Full text of "Proceedings : International Workshop on Subalpine Stone Pines and Their Environment: the Status of Our Knowledge, St. Moritz, Switzerland, September 5-11, 1992"

See other formats

Historic, Archive Document

Do not assume content reflects current
scientific knowledge, policies, or practices.

United^tes
Department
Agriculture

Forest Service

Intermountain
Research Station

General Technical
Report INT-GTR-309

July 1994

Proceedings — International
Workshop on Subalpine
Stone Pines and Their
Environment: the Status
of Our Knowledge

Statements by contributors from outside the U.S. Department
of Agriculture are solely the views of the authors and do not re-
flect policies or opinions of the Department.

The use of trade or firm names in this publication is for reader
information and does not imply endorsement of any product or
service by the U.S. Department of Agriculture or other organi-
zations represented here.

Intermountain Research Station
324 25th Street
Ogden, UT 84401

Proceedings — International
Workshop on Subalpine Stone
Pines and Their Environment:

the Status of Our Knowledge

St. Moritz, Switzerland, September 5-11, 1992

Compilers:

Wyman C. Schmidt
Intermountain Research Station
Forest Service

U.S. Department of Agriculture

Friedrich-Karl Holtmeier
Department of Geography
University of Munster
Germany

Workshop Sponsors:

intermountain Research Station
Westfalische Wilhelms-Universitat,

Landscape Ecology Division, Germany
Swiss Federal Institute for Forest, Snow,

and Landscape Research
Karl and Sophie Binding Foundation,

Switzerland

FOREWORD

This proceedings is a product of the first comprehensive
and international examination of stone pine species of the
world. It reports the papers that were presented at the Inter-
national Workshop on "Subalpine Stone Pines and Their
Environment — The Status of Our Knowledge" held
September 5-12, 1992, at St. Moritz, Switzerland.

It is generally recognized that there are five subalpine
stone pines in the world— all in the Northern Hemisphere-
one in North America and four in Europe and Asia. Ail share
many of the same characteristics and occupy similar ecologi-
cal niches. Stone pines are defined as those species of Pinus
that have five-needled fascicles, have wingless seeds, and
have cones that remain closed at maturity.

How did this international workshop come about? Comple-
tion of a whitebark pine workshop in 1987 and a major sym-
posium in 1989 in Bozeman, MT, U.S.A., sparked interest in
high-elevation ecosystems of western North America. One
section of this popular symposium specifically addressed high-
elevation ecosystems of the world. Information presented
there hinted that we could learn much about whitebark pine
ecosystem analogs from other parts of the world that have
similar physical and biological features. In addition many of
the Eurasian forests have a long history of human influence.
The logical place to look for this type of information was the
world's forest areas that supported subalpine stone pines.
A cursory examination of the five commonly recognized sub-
alpine stone pines indicated they had many similarities.

With this as a background, a small group of western U.S.
researchers and managers held a planning session with
Dr. Holtmeier from Muenster, Germany, in Missoula, MT, in
early 1990. This planning spawned the notion that it was
time that we started doing a better job of collaborating be-
tween colleagues in Europe, Asia, and North America. The
logical starting point was an international workshop that ad-
dressed those items that held common interest for European,
Asian, and North American specialists. So, the international
planning transmission was put into gear and 2V2 years later a
workshop was held in St. Moritz, Switzerland.

The primary objectives of the Stone Pine Workshop were:

1 . To improve international cooperation and collaboration
between scientists.

2. To exchange research findings.

3. To determine knowledge gaps in the ecology and man-
agement of subalpine stone pines of the world.

All five subalpine stone pines were covered by authors of
48 papers from Europe, Asia, and North America. They were:

Species

Swiss stone pine
{Pinus cembra)

Siberian stone pine
{Pinus sibirica)

Natural range

European Alps, Carpathian
Mountains

Siberian Russia and Northern
Mongolia (95 percent in
Russia, 5 percent in Mongolia)

Japanese stone pine Japan, Korea, Eastern Siberia
{Pinus pumila)

Korean stone pine Northeastern China, North Korea,
{Pinus koraiensis) Honshu Japan, southeastern

Russia

Whitebark pine
{Pinus albicaulis)

Western North America

The 1 2 countries represented and the number of papers
from each country were: Russia (9), China (1), Japan
(4), Switzerland (5), Austria (4), Italy (3), France (1), Yugosla-
via (1), Czechoslovakia (1), Romania (1), Germany (3), and
United States (15). The main subjects covered were:

• Evolution and taxonomy.

• Basic ecology of stone pine species.

• Growth characteristics.

• Influence of environmental factors.

• Regeneration.

• Importance to wildlife.

• Forest structure and dynamics.

• Forest management.

• Research needs.

Stone pines are critical forest components on vast areas.
Pinus sibirica alone covers 40 million hectares, mainly in
Siberia, and is a key component of the economy and biology.
P. l<oraiensis accounts for one-fourth of the timber production
in China. P. albicaulis is key to the survival of wildlife such as
the grizzly bear and is a protector of high-elevation water-
sheds and visual resources. P. cembra plays a key role in
avalanche protection, wood production, and recreation in
Europe. P. pumila is a key high-elevation species for stabili-
zation of mountain slopes in Japan and Kamchatka. These
species have many things in common, likely starting with a
common ancestor, probably P. sibirica. They also share
common methods of seed dispersal with the North American
and Eurasian nutcrackers. Natural selection processes have
resulted in a real mutualism between subalpine stone pines
and the nutcrackers.

This workshop was sponsored by the Intermountain
Research Station, USDA Forest Service, U.S.A.; the
Westfalische Wilhelms-Universitat, Landscape Ecology Divi-
sion, Germany; the Swiss Federal Institute for Forest, Snow,
and Landscape Research, Switzerland; and the Karl and
Sophie Binding Foundation, Switzerland.

We want to acknowledge the people and the organizations
that supported them for their enthusiastic and dedicated ef-
forts in the planning and conduct of this international work-
shop. Three field tours were conducted by the hosts of the
workshop. The first tour entitled "Engadine Cembra Pine-
Larch Forest," was in a cembra pine-larch forest overlooking
St. Moritz, a heavily used recreation area in the Engadine.
The primary subjects were tree regeneration and succession
in avalanche tracks and, at the upper timberline, squirrel and
nutcracker interactions, human historical influence, grazing
and game, and the pressures of modern tourism on the for-
ests and landscape.

The second tour, entitled "Ecology and Technique of High-
Altitude Afforestation," dealt primarily with management of
avalanche-prone areas. It was conducted in the Stillberg Ex-
perimental Area near timberline in the Dischma Valley near
Davos. It demonstrated the use and design of physical barri-
ers for avalanche control and reforestation methods with
cembra pine, larch, and mugo pine for long-term reduction of
avalanche problems.

The third tour, entitled "Subalpine and Alpine Environ-
ments," emphasized vegetation succession, particularly of
cembra pine and larch following the gradual retreat of the
Morteratsch Glacier. Also included in this tour was a visit to
the glacier areas at Diavolezza on the Bernina Pass where
recreation values are paramount. From the aerial tram to the

mountain station (2,973 m) we viewed the strong contrasts of
vegetation and glaciers on the north slopes of the Bernina
Pass area.

For all the hosts of the enlightening field tours and arrange-
ments at St. fvloritz we owe a special thanks. Those included
Drs. W. Schonenberger, R. Hasler, J. Senn, R. Lassig,
F.-K. Holtmeier, and Herr U. Wasem — also Engadine
Foresters J. Altman, I. A. Bisaz, and G. Bott. We also thank
Prof. Dr. H. Steinlin from the Binding-Siftung who attended
the workshop as representative of the Binding Foundation.
All of these people and their staffs were wonderful hosts who
contributed so much to the success of this first international
conference on stone pines.

A special thanks goes to Kathy McDonald, Intermountain
Research Station, U.S.A., who processed all the manuscripts.

The organizing committee was composed of the following
people:

Dr. W. Schonenberger
Dr. R. Hasler
Dr. J. Senn
Swiss Federal Institute for
Forest, Snow and Landscape Research
Zurcher Str. 1 1 1
CH-8903 Birmensdorf
Switzerland

Prof. Dr. F.-K. Holtmeier

University of f^unster
Department of Geography
Landscape Ecology Division
Robert-Koch-Str. 26
W-4400 Munster
Germany

Dr. Wyman C. Schmidt
USDA Forest Service
Intermountain Research Station
Forestry Sciences Laboratory
Montana State University
Bozeman, MT, 59717-0278, U.S.A.

Drs. Holtmeier and Schmidt served as technical evaluators
of all the papers.

As a consequence of this international workshop, a resolu-
tion was passed by the scientists in attendance that aimed to-
ward greater coordination and sharing of research findings.
The resolution was:

1 . To organize an International Association for Stone Pine
Research.

2. To form an international organizing committee by elec-
tion that could probably meet in Tomsk, Russia, in 1993 (or
later) to (a) establish the objectives of the Association; and
(b) establish the organizational structure of the Association.

3. The organizing committee to develop the preliminary
plan for the next International Stone Pine Workshop.

4. To hold the next International Stone Pine Workshop at
Tomsk, Russia, in 1995 (or later).

The world faces many environmental and resource chal-
lenges. Many of these challenges are in temperate forests of
the Northern Hemisphere. In these, stone pines are a signifi-
cant forest component and one in which our knowledge base
is certainly not complete. Sharing the knowledge developed
from similar ecosystems and building on that solid base is
key to a better understanding of how our subalpine forests
function ecologically. We can learn much from each other.
The international workshop that generated this proceedings is
a good first step in that direction. Let us do what we can to-
gether to keep the momentum.

Wyman C, Schmidt F.-Karl Holtmeier

VORWORT

Bei den Arten handelte es sich um:

Dieser Band ist eine erste umfassende internationale Studie
iiber die subalpinen Steinkiefern der Erde. Er enthalt die
Vortrage, die anipiich des Internationalen Workshops "Subal-
plne Stone Pines and Their Environment — The Status of Our
Knowledge" gehalten wurden. Dieser Workshop fand vom
5. bis 1 1 . Septennber 1 992 in St. Moritz (Schweiz) statt. Die
funf subalpinen Steinkieferarten sind ausschliepiich in der
Nordhemisphare verbreitet — davon eine in Nordamerika
und vier in Eurasien. Diese Arten weisen viele gemeinsame
Merkmale auf und besetzen auch uhnliche okologische
Nischen. Unter Steinkiefern versteht man die Arten des
Genus Pinus mit funfnadeligen Kurztrieben, flugellosen
Samen und Zapfen, die im Reifezustand geschlossen
bleiben.

Die Vorgeschichte dieses Internationalen Workshops
beginnt 1 987. Damals fand in Bozeman ein Workshop uber
die Whitebark Pine {Pinus albicaulis) statt, dem 1989 ein
gropes Symposium, ebenfalls in Bozeman, folgte. Beide
Veranstaltungen stiepen auf eine grope Resonanz und fuhrten
zu einem wachsenden Interesse an den Hochgebirgsoko-
systemen des westlichen Nordamerika. Eine Vortragssektion
dieses Symposiums war den Hochgebirgsokosystemen der
Welt gewidmet. Dabei zeigte sich, dap aus einem Vergleich
mit Steinkiefer-Okosystemen anderer in ihrer Naturaus-
stattung mit dem nordamerikanischen Verbreitungsareal der
Whitebark Pine vergleichbarer Regionen der Erde auch
wichtige Informationen zum besseren Verstandnis der
Whitebark Pine - Walder gewonnen werden konnen. Als ein
wesentlicher Unterschied zu den amerikanischen Whitebark
Pine - Waldern erwies sich die zum Tell schon einige tausend
Jahren zuruckreichende anthropogene Beeinflussung der
eurasiatischen Steinkiefernwalder durch den Menschen. Mit
Sicherheit lassen sich aus den dortigen Verhaltnissen auch
nutzliche Hinweise auch potentielle Auswirkungen men-
schlicher Eingriffe in den amerikanischen Whitebark Pine -
Waldern gewinnen.

Anfang 1990 traf sich dann eine kleine Gruppe von Wissen-
schaftlern und Forstleuten aus den westlichen Vereinigten
Staaten mit Dr. Holtmeier (Munster, Deutschland) in Missoula
(Montana), um uber weiterfuhrende Forschungen iiber die
subalpinen Steinkiefern zu beraten. Der nachste Schritt
sollte eine Verbesserung und Intensivierung der wissen-
schaftlichen Zusammenarbeit der Fachleute aus Europa,
Asien und Nordamerika sein. Man kam uberein, ein en
Internationalen Workshop zu veranstalten, mit dem Ziel, dort
Fragen von gemeinsamem Interesse zu diskutieren und neue
Forschungsansatze zu entwickeln. Unmittelbar nach diesem
Treffen begannen die Vorbereitungen zu diesem Workshop,
der dann zweieinhalb Jahre spater in St. Moritz (Schweiz)
stattfand.

Seine Hauptziele waren:

1 . Verbesserung und Intensivierung der internationalen
Zusammenarbeit.

2. Austausch von Forschungsergebnissen.

3. Forschungsbedarf hinsichtlich der Okologie und
Bewirtschaftung subalpiner Steinkiefernwalder aufzuzeigen.

Uber alle funf Steinkieferarten wurden Vortrage gehalten.
Insgesamt waren es 48 Referate, an denen Wissenschaftler
aus 12 Staaten beteiligt waren.

Art

Arve (Synonym: Zirbe)
{Pinus cembra)

Sibirische Arve
{Pinus sibirica)

Sibirische Zwergarve
(Japanische
Steinkiefer)
{Pinus pumila)

Koreanische Steinkiefer
{Pinus lioraiensis)

Weiprinden-Kiefer
{Pinus albicaulis)

Naturliches
Verbreitunggebiet

Europaische Alpen, Karpaten

Russisch Sibirien und ndrdliche
Mongolei (95 percent in
Rupiand; 5 percent in der
Mongolei)

Japan, Korea, Ostsibirien

Nordost-China, Nordkorea,
Honshu, Siidostasien

Westliches Nordamerika

Folgende Lander waren vertreten (Zahl der Beitrage in
Klammern): Rupiand (9), China (1), Japan (4), Schweiz (5),
Osterreich (4), Italien (3), Frankreich (1), Yugoslawien (1),
Tschechoslowakei (1), Rumanien (1), Deutschland (3) und
USA (15)

Die Hauptthemen waren:

• Evolution und Taxonomie.

• Autokologie der subalpinen Steinkieferarten.

• Wachstumscharakteristik.

• Einflup der Standortfaktoren.

• Regeneration.

• Bedeutung der Steinkiefern fiir die Tiere.

• Waldstrukturen und -dynamik.

• Waldbewirtschaftung und -pflege.

• Forschungsbedarf.

Die Steinkiefern stellen einen hohen Anteil an der Wald-
bedeckung groper Gebiete. Allein Pinus sibirica bedeckt ein
Areal von 40 Millionen Hektar. Der gropte Teil davon liegt in
Sibirien, wo der Sibirischen Arve sowohl in ihrer Bedeutung
fur die Wirtschaft als auch im Hinblick auf die biologische
Situation die Rolle eines Schlusselfaktors zukommt. Pinus
koraiensis stellt ein Viertel der Holzproduktion in China.
Pinus albicaulis ist unter anderem fur den Grizzly-Bar von
existentieller Bedeutung. Auperdem erfullt diese Art in hoch-
gelegenen Wassereinzugsgebieten eine wichtige Schutz-
funktion und erhoht zudem wesentlich den asthetischen Reiz
der Gebirgslandschaft. Pinus cembra kommt in Europa eine
wichtige Rolle bei der Lawinenvorbeugung durch Schutz-
walder zu. Zugleich liefert sie ein fiir Tafelungen und Schnitz-
erein geschatztes Holz. Auch sie ist von hohem asthetischen
Wert. Pinus pumila tragt in Japan und Kamtschatka ent-
scheidend zur Stabilisierung steiler Gebirgshange bei. Die
vielen Gemeinsamkeiten der genannten Arten haben ihre
Ursache vermutlich in ihrer Abstammung von einer Art,
moglichenweise von Pinus sibirica. Charakteristisch ist fiir sie
auch die Verbreitung ihrer Samen durch den nordamerika-
nischen und europaischen Tannenhaher. Zwischen den
subalpinen Steinkiefern und den Hahern hat sich im Laufe
der Evolution ein echter Mutualismus entwickelt.

Der Workshop wurde unterstutzt durch die Intermountain
Research Station, USDA Forest Service, USA, die West-
falische Wilhelms-Universitat (MOnster, Deutschland), die
Eidgenossische Forschungsanstait fur Wald, Schnee und
Landschaft (Schweiz) sowie durch die Karl und Sophie Bind-
ing - Stiftung.

Wir danken alien Personen und Organisationen, die mit
ihrem Enthusiasmus und ihrer tatkraftigen Unterutzung die
Planung und die Durchfuhrung dieses Workshops ermbglicht
haben.

Wahrend des Workshops fanden unter der Leitung der
Gastgeber drei Exkursionen statt. Die erste fuhrte in die
Larchen-Arvenwalder oberhalb des bekannten Engadiner
Hohenkurortes St. Moritz. Hauptthemen waren die Ver-
jungung und Sukzession in Lawinenzugen und im Wald-
grenzbereich, die Einbindung von Eichhornchen und Tannen-
haher in das Okosystem des Larchen-Arvenwaldes, die
Einflusse von Beweidung und Wild sowie die Belastungen
der Walder und der Landschaft durch den modernen
Tourismus.

Die zweite Exkursion, die sich im wesentlichen mit okologi-
schen Problemen und Techniken bei der Hochlagenauf-
forstung und der Behandlung von Lawineneinzugsgebieten
befapte, fuhrte in die im Waldgrenzbereich gelegenen Ver-
suchsflachen am Stillberg im Dischmatal bei Davos. Dort
wurden kunstliche Lawinenverbauungen und Wieder-
aufforstungen mit Arven, Larche und Legfohre als Mittel zur
Lawinenvorbeugung demonstriert.

Die dritte Tour, die unter dem Thema "subalpine und alpine
Lebensraume" stand, fuhrte zunachst in das Morteratsch-Tal.
Dort ging es vor allem um die Vegetationssukzession im
Vorfeld des allmahlich zuruckweichenden Morteratsch-
Gletschers unter besonderer Berucksichtigung der
Besiedlung durch Arve und Larche. Mit einem Besuch der
Diavolezza am Berninapap fand die Exkursion ihren
Abschlup. Von der Bergstation (2973 m) der Luftseilbahn bot
sich ein uberwaltigender Blick auf die stark vergletscherte
Nordseite der Berninagruppe.

Fur alles, die eriebnisreichen und lehrrreichen Exkur-
sionen sowie die Veranstaltungen in St. Moritz, schulden wir
den Gastgebern — Drs. W. Schonenberger, R. Hasler,
J. Senn, R. Lassig, F.-K. Holtmeier und Herr U. Wasem —
sowie den Engadiner Fbrstern J. Altmann, I. A. Bisaz, und
G. Bott besonderen Dank. Nicht zuletzt geht unser Dank
auch an Herrn Prof. Dr. H. Steinlin, der als Vertreter der
Binding-Stiftung selbst an dem Workshop teilnahm. Sie alle
haben mit ihrem gro^en Engagement bei der Planung und
der Durchfuhrung zum Gelingen dieses ersten internationalen
Workshops uber die Steinkiefern beigetragen.

Besonderer Dank gilt auch Frau Kathy McDonald und
Herrn Richard Klade (Intermountain Research Station, USA)
fur die redaktionelle Bearbeitung der Manuskripte.

Das Organistaionskomitee setzte sich zusammen aus:

Dr. W. Schonenberger
Dr. W. Hasler
Dr. J. Senn
Eidgenossische Forschungsanstait
fur Wald, Schnee und Landschaft
Zurcher Strape 1 1 1
Ch-8093 Birmensdorf
Schweiz

Prof. Dr. F.-K. Holtmeier
Westfalische Wilhelms-Universitat
Institut fur Geographie
Abteilung Landschaftsokologie
Robert-Koch-Strape 29
D-48149 Munster
Deutschland

Dr. Wyman C. Schmidt
USDA Forest Service
Intermountain Research Station
Forestry Sciences Laboratory
Montana State University
Bozeman, Montana, USA 59717-0278

Die Vorbereitung der Manuskripte fur die redaktionelle
Bearbeitung wurde von Dr. Holtmeier und Dr. Schmidt
durchgefuhrt.

Wahrend des Workshops veranschiedeten die Teilnehmer
eine Resolution, die eine noch weitergehende Koordination
der Forschung und bessere gegenseitige Information uber
neue Forschungsergebnisse zum Ziel hat. Sie umfapt
folgende Punkte:

1 . Grundung einer Internationalen Vereinigung fur die
Erforschung der Steinkiefern.

2. Bildung eines Internationalen Organisationskomitees,
das erstmals 1993 (oder spater) in Tomsk (Rupiand)
zusammentreten soli, um die kunftigen Ziele dieser Ver-
einigung zu formulieren und eine Organisationstruktur zu
schaffen.

3. Planung des nachsten Internationalen Workshops uber
die Steinkiefern.

4. Durchfuhrung dieses Workshops voraussichtlich 1995
(Oder spater) in Tomsk.

Die Welt sieht sich vielen Herausforderungen durch
Umweltprobleme und Ressourcennutzung gegenuber.
Davon sind auch die Gebirgswalder und borealen Walder der
Nordhemisphare betroffen, zu denen auch die Steinkiefern-
walder gehoren. Das Wissen uber sie ist noch luckenhaft.
Indem wir unser Wissen zusammentragen bilden wir eine
solide Grundlage, von der aus es leichter sein wird, tiefere
Einsichten in die okologischen Wirkungsgefuge dieser
Waldokosysteme zu gewinnen. Der Internationale Workshop
in St. Moritz mit den hier zusammengestellten Beitragen ist
ein erster Schritt in diese Richtung. Jetzt geht es darum,
diesen Weg weiterzugehen.

Wyman C. Schmidt F.-Kari Holtmeier

CONTENTS

AUTHORS TITLE PAGE

Wyman C. Schmidt Distribution of Stone Pines 1

FIELD TOURS

Swiss Hosts Tour Descriptions 8

Mihailo Grbic

Friedrich-Karl Holtmeier Introduction to the Upper Engadine and Its Forest 9

EVOLUTION AND TAXONOMY
Konstantin V. Krutovskli Genetic Differentiation and Phylogeny of Stone Pine Species Based on

Dmitri V. Politov Isozyme Loci 19

Yuri P. Altukhov

Hermann Mattes Coevolutional Aspects of Stone Pines and Nutcrackers 31

Dmitri V. Politov Allozyme Polymorphism, Heterozygosity, and Mating System of Stone Pines 36

Konstantin V. Krutovskii

Diana F. Tomback Genetic Population Structure and Growth Form Distribution in Bird-Dispersed Pines 43

William S. F. Schuster

BASIC ECOLOGY

Ernst Frehner Experiences With Reproduction of Cembra Pine 52

Walter Schonenberger

P. L. Gorchakovsky Distribution and Ecology of Siberian Stone Pine in the Urals 56

Rudolf Hasler Ecophysiological Investigations on Cembran Pine at Timberline in the Alps, An Overview 61

Peter A. Khomentovsky A Pattern of Pinus pumila Seed Production Ecology in the Mountains of Central

Kamtchatka 67

Takayoshi Koike Needle Longevity and Photosynthetic Performance in Cembran Pine and Norway Spruce

Rudolf Hasler Growing on the North- and East-Facing Slopes at the Timberline of Stillberg in the

Hans Item Swiss Alps 78

Luo Ju Chun The Broad-leaved Korean Pine Forest in China 81

Tad Weaver Climates Where Stone Pines Grow, A Comparison 85

GROWTH CHARACTERISTICS

Kurt Holzer Growth of Swiss Stone Pines That Originated From and Were Planted at Several

Altitudes in the Austrian Alps 91

Takuya Kajimoto Seasonal Patterns of Growth and Photosynthetic Activity of Pinus pumila Growing on the

Kiso Mountain Range, Central Japan 93

Herbert Kronfuss Height Growth in Cembran Pine as a Factor of Air Temperature 99

Josef Senn Survival and Growth of Planted Cembran Pines at the Alpine Timberline 105

Walter Schbnenberg
Ueli Wasem

AUTHORS TITLE PAGE

INFLUENCE OF ENVIRONMENTAL FACTORS

Mihailo Grbic Changes of Swiss Stone Pine Aphid Life Cycle, Density, and Population Structure in

High-Altitude Swiss Stone Pine Afforestation 111

Raymond J. Hoff Genetic Consequences and Research Challenges of Blister Rust in Whitebark Pine

Susan K. Hagle Forests 118

Richard G. Krebill

Karel Kahik Performance of Pinus cembra, P. peace, and P. strobiformis Within Air-Polluted Areas 127

Todd Kipfer Competition and Crown Characteristics of Whitebark Pine Following Logging in

Katherine Hansen Montana, U.S.A 130

Ward McCaughey

Penelope Morgan Fire Ecology of Whitebark Pine Forests of the Northern Rocky Mountains, U.S.A 136

Stephen C. Bunting
Robert E. Keane
Stephen F. Arno

Tad Weaver Vegetation Distribution and Production in Rocky Mountain Climates — With Emphasis

on Whitebark Pine 142

REGENERATION

I. Blada Variation in Size and Weight of Cones and Seeds in Four Natural Populations of

N. Popescu Carpathian Stone Pine 154

Mitsuhiro Hayashida Role of Nutcrackers on Seed Dispersal and Establishment of Pinus pumila and

P. pentaphylla 1 59

Harry E. Hutchins Role of Various Animals in Dispersal and Establishment of Whitebark Pine in the

Rocky Mountains, U.S.A 163

Hermann Mattes Size of Pine Areas in Relation to Seed Dispersal 172

Ward W. McCaughey The Regeneration Process of Whitebark Pine 179

Ikuko Nakashinden Japanese Stone Pine Cone Production Estimated From Cone Scars, Mount

Kisokomagatake, Central Japanese Alps 188

Diana F. Tomback Effects of Seed Dispersal by Clark's Nutcracker on Early Postfire Regeneration of

Whitebark Pine 193

Vladislav N. Vorobjev New Trend in Dendrochronology: 1. Theoretical Principles of Reprochronology 199

V. N. Vorobjev New Trend in Dendrochronology: 2. Method of Retrospective Study of Seminiference

S. N. Goroshkevich Dynamics in Pinaceae 201

D. A. Savchuk

IMPORTANCE TO WILDLIFE

Ronald M. Lanner Nutritive Value of Whitebark Pine Seeds, and the Question of Their Variable Dormancy 206

Barrie K. Gilbert

David J. Mattson Bear Use of Whitebark Pine Seeds in North America 212

Daniel P. Reinhart

Diana F. Tomback Ecological Relationship Between Clark's Nutcracker and Four Wingless-Seed Strobus

Pines of Western North America 221

AUTHORS TITLE PAGE

FOREST STRUCTURE AND DYNAMICS

Giovanna De Mas Structure of Swiss Stone Pine Stands in Northeastern Italy 226

Elena Piutti

S. Yu. Grishin Role of Pinus pumila in Primary Succession on the Lava Flows of Volcanoes of

Kamchatka 240

Robert E. Keane Decline of Whitebark Pine in the Bob Marshall Wilderness Complex of Montana, U.S.A 245

Penelope Morgan

Renzo Motta Some Aspects of Cembran Pine Regeneration in the Italian Cottian Alps 254

Alberto Dotta

Pietro Piussi Mixed Cembran Pine Stands on the Southern Slope of the Eastern Alps 261

Siegfried Sauermoser Current Distribution of Cembra Pine in the Lechtal Alps 269

E. P. Smolonogov Geographical Differentiation and Dynamics of Siberian Stone Pine Forests in Eurasia 275

Use WI6rick The Cembran Pine in the French Alps: Stand Dynamics of a Cembran Pine Forest

in Tueda (Savoy, France) 280

FOREST MANAGEMENT

Werner Frey Silvicultural Treatment and Avalanche Protection of Swiss Stone Pine Forests 290

Riet Gordon Importance and Silvicultural Treatment of Stone Pine in the Upper Engadine (Grisons) 294

Jorg Heumader Cultivation of Cembran Pine Plants for High-Elevation Afforestations 298

Katherine C. Kendall Whitebark Pine Conservation in North American National Parks 302

RESEARCH NEEDS

Wyman C. Schmidt Research Needs in Whitebark Pine Ecosystems 309

Vladislav N.Vorobjev Problems of Comprehensive Investigation, Utilization, and Reproduction of

Nina A. Vorobjeva Russian Cedar Pine Forests 314

DISTRIBUTION OF STONE PINES

Wyman C. Schmidt

The five stone pine species that are the subject of this
proceedings occur only in the Northern Hemisphere — one
species in North America and the other four in Europe
and Asia. Taxonomically they fall in section Strobus and
subsection Cembrae. They are characterized by having
five-needled fascicles, cones that do not open at maturity,
and seeds that are wingless.

The distributions of the five species are reasonably well
defined, but updates on their distribution usually show
minor changes. Also, in some cases, such as with white-
bark pine (Pinus albicaulis), the species occurs only at
high-elevation sites, and there may be long distances be-
tween the mountain ranges that support this species. This
species ocoirs over a large geographic area; however, there
are usually large intervening valleys and lowlands not
suitable for whitebark pine forests. The same is true for
the other stone pines. Therefore, just because a distribu-
tion boundary of a species encompasses a large area, it
does not mean that all the mountains and valleys are
fully occupied by that species.

For the purposes of this proceedings, we have chosen to
show the general distribution on the maps of the five spe-
cies and do not necessarily include all of the individual
outliers.

The distribution maps that follow are adapted fi'om those
pubUshed by Critchfield and Little (1966). The maps show
that the five species are mostly discrete populations, but
there are overlapping distributions of Siberian stone pine
(Pinus sibirica) and Japanese stone pine (P. pumila), and
an overlap of Japanese stone pine and Korean stone pine
(P. koraiensis).

Whitebark pine occurs only in North America at high
elevations fi*om central British Columbia south through
the Cascade Mountains of Washington and Oregon to the
southern Sierra Nevada of California, in a limited area of
V Nevada, and extensively in Montana, Idaho, and western
Wyoming. It occurs just below the timberline and extends
downward elevationally into associations with several
other conifers.

Swiss stone pine grows at high elevations in the Alps and
Carpathian Moimtains. It occurs mostly in Switzerland,
Austria, northern Italy, southeastern France, Romania,
Czechoslovakia, and Poland. Although Swiss stone pine
is considered closely related to Siberian stone pine, over
2,000 kilometers presently separate the two species.

Swiss stone pine commonly grows in association with
other conifers.

Siberian stone pine is widely distributed in Eurasia fi"om
the Ural Moimtains through western and central Siberia to
northern MongoUa — an east-west distance of about 4,000
kilometers. It overlaps the range of Japanese stone pine
on the eastern portion of its distribution near Lake Baikal
in Siberia. Different than both whitebark pine and Swiss
stone pine, which only occupy high-elevation sites, Sibe-
rian stone pine (usually called cedar-pine or Siberian ce-
dar by Russians) commonly grows in the plains and river
valleys, as well as in the mountain areas.

Japanese stone pine grows in northeastern Siberia and
in subalpine parts of Japan, Korea, and Manchvuia. It ex-
tends north almost to the Arctic Ocean (70° N.), west to
Mongolia and to the Lake Baikal area, and south to Korea
and Honshu, Japan. It grows from low elevations in the
north to high elevations in the south. It commonly forms
thickets just below barren tundra on tops of mountains or
lowlands in the extreme north. Like Siberian stone pine,
it extends nearly 4,000 kilometers fi*om east to west and
3,000 kilometers north to south.

Korean stone pine occurs throughout Korea and eastern
Manchuria into southeastern Siberia as well as on the is-
lands of Honshu and Shikoku in Japan. It tends to grow
in more maritime conditions and lower moimtains than its
neighbor — the Japanese stone pine. Its geographic range
overlaps that of the Japanese stone pine, but the species
occupy different ecological niches. Korean stone pine, im-
like its counterparts, will grow in association with hard-
woods where near-maritime cUmates prevail. Its valuable
wood makes it a much sought after species.

The following maps illustrate the geographic distribu-
tion of the five stone pines. These descriptions are based
primarily on those of Critchfield and Little (1966) and
Mirov (1967).

References

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis-
tribution of the pines of the world. Misc. Publ. 991.
Washington, DC: U.S. Department of Agricvdture, For-
est Service. 97 p.

Mirov, N. T. 1967. The genus Pinus. New York: Ronald
Press Company. 602 p.

NORTH AMERICA

135° 130° 125° 120° 115° 110°

130° 125° 120° 115° 110° 105° 100°

EUROPE

10° 5° 0° 5° 10° 15° 20° 25° 30° 35°

0° 5° 10° 15° 20° 25° 30° 35°

ASIA

Field Tours

International Workshop
St. Moritz 1 992

Field tours were an important component of the Stone Pine Workshop. First,
the tours provided an informal setting for scientists from the 12 countries to become
acquainted with each other and their research; second, the tours demonstrated the
research and management of Swiss stone pine forests in the rugged and spectacularly
beautiful mountains of Switzerland.

The following section is the handout from the Swiss hosts used to describe
the three major field tours. The drawings by a gifted scientist, Mihailo Grbic, from
Belgrade, Yugoslavia, depict a symbolic participant, Dr. Stone Pine, enjoying the rather
strenuous excursions. Grbic also created the welcome drawing used as sections
division indicators in this publication.

TOUR DESCRIPTIONS

Swiss Hosts

We will try to use pleasant weather for the field tours.
Therefore this preliminary schedule may change a little.
Usually weather is fine in September. However, snow and
fi-ost are possible. Do not forget warm clothing, tough foot-
wear, backpack, sunblock, and simglasses. During the field
toiu-s we will have lunch fi-om the backpack.

Engadine Cembra Pine-Larch Forest

Sunday 6th. St. Moritz, Muottas da Schlarigna. Walk
fi-om St. Moritz Bad (Hotel "Laudinella'')-Stazer Wald-
Muottas da Schlarigna-St. Moritz Bad (1,800 to 2,300 m
above sea level [a.s.l.]).

Cembra pine-larch forest: composition, structure, growth
characteristics, old cembra pines, regeneration pattern,
understory vegetation, avalanche tracks with green alder
{Alnus viridis) formation, influence of seed-caching ani-
mals (squirrel, nutcracker), effects of seed dispersal by
the thick-billed European nutcracker (Nucifraga c.
caryocatactes).

Human influence on the forest distribution pattern,
human-caused timberline, influence of grazing livestock
and game, and of toiu"ism; invasion of rarely used and
abandoned alpine and forest pastures by trees, influence
of microtopography and microclimate on alpine vegetation
and on growth of invading trees.

Ecology and Technique of High-
Altitude Afforestation

Tuesday 8th. Davos, Stillberg. Bus fi"om St. Moritz-
Albula pass-Davos-Teufi (Dischma Valley, 1,700 m),
5-hour walk to and through Stillberg experimental area
near timberline at 2,000 to 2,230 m a.s.l. and back to the
valley bottom (Teufi).

The experimental area lies on a steep avalanche catch-
ment slope. Construction of avalanche barriers. Dam-
ages, diseases, mortality and development of a 17-year-old

plantation of Pinus cembra, Larix decidua, and P. mugo.
Influence of different site factors such as wind, tempera-
ture, irradiation, snow movement, duration of snow cover,
soil properties, vegetation on plantation success. Eco-
physiology of subalpine timberline trees. Conclusions for
plantation technique and spatial patterns.

Subalpine and Alpine Environments

Friday 11th. Morteratsch Glacier, Diavolezza.
Morteratsch Glacier (2,000 m, 2- to 3-hour walk). Train to
Morteratsch station, walk to Morteratsch Glacier via Alp
Chiinetta (2,052 m); old-growth cembra pine forests, re-
generation, invasion of the lateral moraine (fi-om 1850)
and the glacier forefield by cembra pine and larch.

Diavolezza (2,973 m). Train fi-om Morteratsch station
through the Bemina Valley to Diavolezza station (2,200 m)
on Bemina Pass, and by ropeway to Diavolezza. Magnifi-
cent view of the Bernina Massive (4,049 m) with Pers Gla-
cier and Morteratsch Glacier.

★ ***★

In addition to providing valuable scientific and manage-
ment information, the introductory paper by Friedrich-
Karl Holtmeier that follows gave background and baseline
data describing the local Swiss stone pine forests that
were the focus of the field tours.

*****

INTRODUCTION TO THE UPPER
ENGADINE AND ITS FOREST

Friedrich-Karl Holtmeier

Abstract — The Engadine is characterised by great mean eleva-
tion (2,400 m) and a relatively continental climate. Cembra pine
{Pinus cembra) and European larch (Larix decidua) are the most
important tree species. The area was already settled in prehis-
toric time. Since then, mountain forest has been influenced by
human impact. During history timberline became lower by about
150-300 m. At present, the upper forest limit is located at about
2,200-2,300 m. Composition and structure of the forest also
changed in that larch could spread at the cost of cembra pine on
all areas that were suitable for grazing. After grazing pressure
has declined natural succession is revived, and, in the long term,
larch will be replaced by cembra pine, if not specially managed.
Natural stands of Norway spruce (Picea ahies) occur on the
northwest-exposed slopes of the main valleys within an area
characterized by higher humidity. Spruce reaches its upper limit
at 2,000 m. Dwarfed mountain pine (Pinus mugo) and green al-
der (Alnus viridis) are most common in avalanche tracks. Green
alder is to be found mainly on moist sites, whereas mountain
pine prevails under dry conditions. Alder also occurs in bogs.

The Engadine — Engiadina in Rhacto-Romsinic lan-
guages — is located in the eastern part of Switzerland and
comprises the uppermost drainage area of the Inn River
(fig. 1). The main valley trends from southwest to north-
east. By an administrative border the Engadine is divided
into Upper and Lower Engadine. However, landscape
physiognomy of both sections is also different. The Lower
Engadine is deeply cut into the mountains, and mountain
slopes are steep and nagged. The Upper Engadine main
valley, on the other hand, is wide and topography is com-
paratively smooth. Between St. Moritz and the Maloja
Pass the valley is characterized by four lakes separated by
alluvial plains. To the southeast the Bemina Valley gives
access to the famous glaciated Bemina Group.

The Engadine is characterized by great mean elevation
(2,400 m). The bottom of the main valley is located at an
altitude of about 1,800 m. The tributary valleys climb up
to about 2,300 m. The highest peaks' elevations are about
4,000 m (Piz Bernina 4,049 m; fig. 1). In view of its great
elevation the Upper Engadine has been called The Roof
of Europe" by the English.

Due to the geographical location in the central Alps
and the high mass-elevation of the Engadine the climate
is rather continental, characterized by a relatively great

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Friedrich-Karl Holtmeier is Professor of Landscape Ecology and Geogra-
phy, Department of Geography, University of Munster, Germany.

amplitude of mean temperature and comparatively low
amount of precipitation (fig. 2). Annual precipitation de-
creases from about 1,400 mm at Maloja Pass, where the
main valley is open to moisture-carrying air ciurents from
the southwest (fig. 3), to less than 1,000 mm at St. Moritz.
At timberline, however, precipitation is much higher than
at the valley bottom, as can be concluded from the data
available for Julier Hospiz on Julier Pass and for Bemina
Hospiz on Bemina Pass. At both stations, anniial precipi-
tation exceeds 1,500 mm.

Due to the continental climate, the upper limits of veg-
etation and cultivation reach relatively high altitudes. In
the Upper Engadine, forest limit is located at about 2,200-
2,300 m. Solitary crippled trees may still be found at and
even above 2,500 m (Holtmeier 1965, 1974).

In the Upper Engadine, artifacts were found that give
evidence of agriculttire during the Bronze Age (about
3,000 years B.P.), while in the Lower Engadine hunters
and shepherds were already present in the Neohthic Age
(4,000 years ago; Conrad 1940). Trading and mining be-
came increasingly important to the economy dvuing his-
toric time. Thus, there was increasing human impact on
the moimtain ecosystems throughout history. Almost no
untouched nature was left, although the kind and the in-
tensity of the himaan disturbances were different.

DEPRESSION OF TIMBERLINE

Due to the great elevation of the area humans could
only use the mountains up to a certain altitude; there the

Figure 1— Location of the Upper Engadine.

3000

2500

2000

1500

1000 -

500 -

Figure 2 — Annual precipitation. Note the de-
crease from Maloja Pass toward St.. Moritz and
Schuls (Lower Engadine).

Figure 4 — Zuoz Village (1 ,700 m). The original
forest was almost totally removed as a conse-
quence of alpine pasturing (upper slope) and
agriculture (artificial terraces above the village).
Only two small stands were left on the steep
north-exposed slopes of the narrow tributary
valleys and one above the village for protection
from avalanches. Photo taken 1 968.

main influence was concentrated in the forest belt not
only on favorable southern exposures but also on all
sections that were fairly accessible — trough shoulders,
terraces, and similar almost-level areas even in north-
exposed locations. During former times cereals were
grown up to elevations of about 2,300 m, favorable topog-
raphy provided. Many of the high-altitude forests were
cleared for pastoral use, while those at lower elevations

gave way to agriculture. The extent of timberline depres-
sion varied in relation to exposure and topography. On
some easily accessible southern exposures of the main
valley, for example, and of the Fex Valley and Fedoz Val-
ley, the forest was totally removed except for sections
characterized by too great steepness (figs. 4, 5). Other for-
ests were destroyed by ore mining, salt works, and char-
coal production, especially during the Middle Ages. On

Figure 3 — View from Lej dais Chods (northwest-
facing slope of the main valley) southwest to-
ward the gap of Maloja Pass.

Figure 5 — View to southeast-exposed slope of the main
valley. In the foreground, Samaden Village (1 ,720 m).
Forest was removed from all gentle topography (trough
shoulders, lower slope area) and reduced to a narrow
belt on the steep trough wall. Photo taken 1969.

Figure 6 — View from Diavolezza trail to
Bernina Pass (2,328 m). Bernina Pass was
covered by larch-cembra pine forest until the
Middle Ages. Then, the forest became a
victim of mining. Until present, the area is
used in summer for grazing cattle.

the Bernina, Abula, and Jiilier Passes (about 2,300 m) the
forest disappeared as a consequence of mining. These ar-
eas v^^ere then used as alpine pastures (fig. 6).

The upper timberUne receded by 150-300 m compared
to its position during the postglacial thermal optimvmi.
That timberline depression was partly due to a general
cooling. However, at least since the Middle Ages the ef-
fects of the deterioration of the climate on subalpine tree
growth have been compounded or even superimposed by
the human influences mentioned earlier.

COMPOSITION AND STRUCTURE

As a consequence of these activities not only did the up-
per timberline become lower by about 150-300 m, but also
species composition and structures of the moimtain for-
ests changed considerably (Auer 1947; Campell 1944;
Holtmeier 1965, 1967). Larch (Larix decidua) and cembra
pine {Pinus cembra) are the most important tree species
in the Upper Engadine forests. These forests belong to
the so-called silicate type (Larici Cembretum, Ellenberg
1978; Larici-Pinetum cembrae, Oberdorfer 1970) of larch-
cembra pine forest, which is common in the crystalline
Central Alps. The present distribution pattern of both
species in the Upper Engadine can be described as fol-
lows: Cembra pine is mainly spread on steep and inacces-
sible slopes and north exposures, whereas larch domi-
nates on easily accessible locations exposed to the south.

The influence of the local topography on the distribu-
tion pattern is most pronovmced in the trough-shaped
Bever Valley and Bernina Valley (fig. 7). Pure larch
stands are confined to the relatively gentle slope areas
(talus cones, screes, boulder fans, etc.) that extend below
the steep and rocky trough walls, while cembra pine pre-
vails on trough walls and also forms the upper timberline.
This pattern is most conspicuous on southern exposures.
On north-exposed slopes cembra pine also occurs on their
lower parts, mixed vnth larch.

This distribution of larch and cembra pine is mainly
caused by human influences, especially by pastoral use.
Since cembra pine impedes the growth of grass and herbs
on the forest floor, while larch does not, cembra pine was
cleared on slopes suitable for grazing and restricted to
steep, inaccessible locations. Moreover, the wood of cem-
bra pine was used for many purposes such as paneling
and making furniture, milk vessels, or carvings. Cembra
pine also is more susceptible to forest fires than larch,
which is protected by a thick corklike bark. As indicated
by charcoal -rich soil horizons, forest fires frequently
occvured in the past. Probably, many of them were
caused by shepherds when they burned alpine pastures
to remove "weeds" such as Rhododendron ferrugineum
and Juniperus nana, and fires ran out of control. Thus,
for many reasons, larch co;ild spread at the cost of cembra
pine, as was also the case in many other central-alpine
valleys. At many places the high-elevation forest is over-
mature and very sensitive to environmental impacts such
as windthrow and snow slides.

CURRENT CHANGES

Due to modem change of economic structure — tourism
has become the main base of the economy in the
Engadine — grazing pressure on the forests declined, and
cembra pine is invading the former pasture forest. Larch,
which is a light-demanding species, is going to be out-
competed by cembra pine, and natural succession ft-om
larch to cembra pine forests, which had been interrupted
by human disturbances for some hundred years, is re-
vived (Holtmeier 1967, 1990, 1993). Larch will gradually
be replaced by cembra pine, if sites are not specially man-
aged by exposing mineral soil or selective cutting of
cembra pine.

Locally, however, invasion of cembra pine in the imder-
story is hampered by mass outbreaks of the larch bud
moth (Zeiraphera diniana). After having defoliated the
larch crowns, the caterpillars will feed on the cembra
pines in the understory and, together with secondary
parasites, cause severe damage to them. Young cembra
pines become crippled or killed.

In forest gaps smaller than the height of the surround-
ing trees, accvunulation of snow and length of v^dnter snow
cover may critically increase (fig. 8), and consequently
seedlings and saplings of cembra pine may be seriously
damaged by snow fungus {Phacidium infestans).

Cembra pine also is invading abandoned or only rarely
used alpine pastures. Seeds are dispersed by the nut-
cracker {Nucifraga caryocatactes), occasionally far beyond
the present forest limit (Holtmeier 1966, 1974; Mattes
1978, 1982). Larch seeds, however, are prevented from
getting into a suitable seedbed by the dense alpine vegeta-
tion. Thus, it is cembra pine that is the pioneer species
on former pasture areas (fig. 9), while larch only occurs
where bare mineral soil is exposed. Survival of the invad-
ing trees mainly depends on microclimates and resulting
site conditions. They considerably deteriorated when the
forest was removed in the past. Favorable effects of the
general warming on tree growth are superimposed by
unsuitable local site conditions (fig. 10). Consequently,
natural reforestation follows the favorable sites while
other sites will remain treeless probably for a long time.

Figure 7 — Distribution of tree species in the Bernina Valley (top) and
Bever Valley (bottom). Both maps have been redrawn from Holtmeier
(1967).

Figure 8 — Big snow masses are accumulated
in the glade. Incoming solar radiation is inter-
cepted by the trees, while energy loss by out-
going long-wave radiation is almost unimpeded
Due to these conditions snowmelt is consider-
ably delayed, and young cembra pines may be
infected by snow fungus.

As to successful germination and survival of seedlings,
seed dispersal by the nutcracker is much more effective
than wind-mediated transport because many seeds are
cached at sites relatively favorable to germination and
seedling growth (not too long covered with snow). On the
other hand, the seed caches usually provide sufficient soil
moisture because they are 2-4 cm under the surface, thus
being protected from direct insolation. Finally, seeds col-
lected by the nutcracker are characterized by a relatively
high germination capacity (Mattes 1978, 1982), and there
is almost no loss of cached seeds by seed-eating predators.

Figure 9 — Cembra pine invading an abandoned
alpine pasture (about 2,250 m) on the northwest-
exposed slope of the main valley. Invasion fol-
lows mainly convex topography, while sites
covered too long with snow and characterized by
snow creep and avalanches remain treeless.

However, due to high density of seedling clusters some
of the trees will become victims of root competition. More-
over, snow accumulation and duration of the snow cover
may gradually be increased by the influence of the grow-
ing trees on windflow near the surface. At such sites,
yoimg cembra pines covered too long by snow may be
killed by snow fungi such as Phacidium infestans
(Holtmeier 1967, 1974, 1990 ).

Natural invasion by trees (cembra pine, larch, vallows,
rowan, and also spruce) can be observed in front of and on
the lateral moraines of the Morteratsch Glacier and Roseg
Glacier (Holtmeier 1965, 1974, 1990; fig. 11). These gla-
ciers have been retreating more or less continuously by
about 30 m per year since the beginning of our century.
Larch alreadj occurs about 20 years after glacier retreat,
while cembra pine ynW follow later. Then, however,
cembra pine invades these areas much more rapidly than
larch. There is some evidence that areas which became
ice-free about 50 to 70 years ago were more rapidly in-
vaded by cembra pine than those from where the glacier
retreated more than 100 years ago. That might be ex-
plained by mesoclimatic conditions improving due to gla-
cier retreat.

1. Situation during the postglacial thermal optimum

forest 1

ecotone 1 alpine

- 2100

- 2100 ijpr

forest climate replaced by the very locally
changing pattern of microclimates , similar
to that above the previous tinfterline

2. Situation at

present

Figure 10 — Change of microclimatic conditions in the
upper Subalpine after the uppermost forest was re-
moved by humans.

1.5 km 1

^Jk cembrapine ^^^i larch ''iV' alpine vegetation

Figure 11 — Invasion of the forefield and lateral moraine of the retreating Morteratsch Glacier by cembra
pine and larch (schematically).

At places, however, distribution patterns of invading
larches and cembra pines very likely depend on soil mois-
ture available, as can be observed in the forefield of
Morteratsch Glacier, for example (fig. 11). Obviously,
relatively dry sites — such as the well-drained upper part
of the lateral moraines or roche moutonnee areas — can be
invaded by cembra pine more easily than by larch, be-
cause seeds of cembra pine are cached by the nutcracker
in small cracks or fine-soil pockets between the boulders
where sufficient moisture is available for germination and
seedling growth. On the other hand, it depends totally on
chance whether wind-mediated larch seeds will get into
such a seedbed. At the valley bottom, conditions appear
to be more favorable for invasion by larch because of
great numbers of open patches of bare mineral soil cre-
ated by extensive accumulation of fine material between
the boiilders. At such sites soil moisture is usually avail-
able, and there is a better chance for larch seeds to reach
a suitable seedbed.

Spruce (Picea abies) also occurs within cembra pine-
larch forest. Natural stands, however, are confined to the
northwest-exposed slope of the main valley between the
village of Sils and the small Lake of Staz (fig. 12). This
area is characterized by increased humidity due to a cloud
belt that usually appears when the Upper Engadine is in-
fluenced by moisture-carrying air currents from the
southwest (Holtmeier 1965, 1966, 1967). The upper dis-
tribution limit of spruce is located at about 2,000 m. Al-
though spruce is missing between Maloja Pass (1,815 m)
and Sils, there is some evidence that this species immi-
grated from the south across Maloja Pass in 6000 to 5500
B.C. (Campell 1944; Kleiber 1974). The gap can probably
be attributed to human influences. On the opposite slope
spruce does not occur, except for planted trees (Schlatter
1935). That might be explained by lower humidity and
lower soil moisture on that slope exposed to the southeast.
Anyway, humans also could have removed shade-giving
spruce (Holtmeier 1967). After deglaciation, cembra pine
invaded the Upper Engadine earlier than spruce (Kleiber

Figure 12 — Distribution of spruce {Picea abies) on the northwest-exposed slope of the main valley (from
Holtmeier 1967, modified).

Figure 13 — Arboreous and dwarfed mountain pine in
bog near Maloja Pass.

1974). Otherwise it would have been outcompeted by the
latter species — at least at lower elevations (Heitz 1975;
Mattes 1978).

KRUMMHOLZ SPECIES

Dwarf pine {Pinus mugo) and green alder (Alnus viri-
dis) form dense and locally extended krummholz stands.
The krummholz growth of both these species is genetically
predetermined and not climatically induced. That makes
them different from climatically trimmed, flagged, and
matlike "krummholz" that forms the forest-alpine timdra
ecotone in the Rocky Moimtains, for example (Holtmeier
1973, 1974, 1981).

Moreover, there is no contiguous Pinus mugo krumm-
holz belt above the high-stemmed forest as is peculiar to
the northern Alps. In the Upper Engadine, both Pinus
mugo and Alnus viridis occur mainly on slopes endan-
gered by avalanches. Their elasticity enables them to
resist avalanches and snow slides better than upright-
growing tree species, which normally will be eliminated
at such sites. Some slopes, where the original forest had
been destroyed by avalanches and fire, were completely

afforested with moimtain pine. Moimtain pine is also
common in bogs where it grows on slightly convex topog-
raphy (mostly rocky outcrops). At such sites the arbore-
ous growth form also occiu-s, as is the case on Maloja Pass
or in the Lake Staz area, for example (fig. 13).

In general, Pinus mugo prevails on limestone and dolo-
mite, while Alnus viridis is common mainly on silicate
material. In view of this distribution, both species are
said to be vicarious plants. In the Upper Engadine, how-
ever, the availability of soil moistiu^e seems to be the fac-
tor controlling the distribution pattern. There, Pinus
mugo is more common on southern exposures, whereas
Alnus viridis is mainly spread on north-facing slopes pro-
viding sufficient soil moistiire and fi'esh soils rich in nutri-
ents. On svmny slopes, however, Alnus viridis is confined
to moist microsites, as alongside rivlets, for example. On
northern exposures Pinus mugo successfully competes
with Alnus viridis at sites relatively dry and poor in nu-
trients (fig. 14).

Birch {Betula pubescens) is another species that is usu-
ally confined to avalanche tracks and bogs. Birch is able
to recover fi-om breakage by thriving basal shoots. There-
fore, multistemmed growth is quite common to this spe-
cies (fig. 15).

At present, the Upper Engadine has a higher forest
cover compared to 150 years ago and earlier. Since then
many afforestations were carried out to close the gaps
caused by historical hiunan impact. Also "exotic" species
were used — such as Picea pungens, Picea engelmannii, Pi-
nus sibirica, and others (Schlatter 1935). It has become
evident, however, that in this area the native Pinus cem-
bra and Larix decidua are the most successful species.

CONCLUDING REMARKS

The Upper Engadine belongs to the most important
natural ranges of cembra pine-larch forests in the Alps.
However, present distribution, composition, and struc-
tures of these forests can only be explained by permanent
utilization that began in prehistoric time. That makes
these forests much different from almost imtouched high-
altitude forests in remote areas of the Rocky Mountains,
for example.

Since tourism has become the main economic base, while
agriculture and pastoral land use declined, utilization of
the forest also changed. In our day it is not timber produc-
tion but establishment and maintenance of a vigorous
shelter forest (protection from avalanches, soil erosion,
torrential washes, etc.) that is the main objective of forest
management. Locally, grazing of the high-altitude forest
by cattle and game may run counter to the objective and
should be reduced as much as possible.

However, the forest also is of great importance for tour-
ism. Thus, larch, which has a high esthetic value because
of its bright autvmin colors, is favored by special manage-
ment practices interrupting natural succession. On the
other hand, lower elevation forest is still used for produc-
tion of merchantable timber. Swiss stone pine is used in
particular for paneling, carvings, and making furniture of
typical local design.

PINUS MUCO

LARIX DECIDUA

9 <*

PINUS CEMBRA

Figure 14 — Distribution of Pinus mugo and AInus viridis on the northeast-exposed slope of the Bernina vaiiey.
Pinus mugo is confined to relatively dry convex topography, while AInus viridis grows on the moist sites (from
Holtmeier 1967).

REFERENCES

Auer, Christian. 1947. Untersuchungen iiber die natiir-
liche Verjiingung der Larche im Arven-Larchenwald
des Oberengadins. Mitteilungen der Schweizerischen
Anstalt fiir das forstliche Versuchswesen. 25: 3-140.

Campell, Eduard. 1944. Der Wald des Oberengadins im
Wandel der Zeiten. Festschrift 124. Jahresversamm-
lung der Schweizerischen Naturforschenden
Gesellschaft: 93-102.

Conrad, H. 1940. Beitrag zur Frage der urgeschichtHchen
Besiedlung des Engadins. Jahresbericht Historisch-
Antiquarische Gesellschaft Graubixndens. 70.
Jahresbericht: 1-40.

Figure 15 — Birch {Betula pubescens) growing
in an avalanche track.

Ellenberg, Heinz. 1978. Vegetation Mitteleuropas mit den
Alpen in okologischer Sicht. 2d ed. Stuttgart: Verlag
Eugen Ulmer. 981 p.

Heitz, Christian. 1975. Vegetationsentwicklung und
Waldgrenzschwankungen des Spatglazials iind Post-
glazials im Oberhalbstein (Graubiinden/Schweiz).
Biindnerwald. 28(2): 45-52.

Holtmeier, Friedrich-Karl. 1965. Die Waldgrenze im
Oberengadin in ihrer physiognomischen und okolo-
gischen DifFerenzierung. University of Bonn, FRG.
163 p. Dissertation.

Holtmeier, Friedrich-Karl. 1966a. Die Malojaschlange
und die Verbreitung der Fichte. Beobachtungen zur
Klimaokologie des Oberengadins. Wetter und Leben.
18: 105-108.

Holtmeier, Friedrich-Karl. 1966b. Die okologische Funk-
tion des Tannenhahers im Zirben-Larchenwald und an
der Waldgrenze im Oberengadin. Journal fur Omi-
thologie. 4: 253-260.

Holtmeier, Friedrich-Karl. 1967. Die Verbreitung der
Holzarten im Oberengadin unter dem Einflui3 des
Menschen und des Lokalklimas. Erdkunde. 21: 249-258.

Holtmeier, Friedrich-Karl. 1973. Geoecological aspects of
timberlines in Northern and Central Europe. Arctic and
Alpine Research. 5(3): 45-54.

Holtmeier, Friedrich-Karl. 1974. Geookologische Beobach-
tungen und Studien an der subarktischen und alpinen
Waldgrenze in vergleichender Sicht (nordliches Fenno-
skandien/Zentralalpen). Erdwissenschaftliche For-
schung. 8. 130 p.

Holtmeier, Friedrich-Karl. 1981. What does the term
"Krummholz" really mean? Observations with special
reference to the Alps and the Colorado Front Range.
Mountain Research and Development. 1(3/4): 253-260.

Holtmeier, Friedrich-Karl. 1990. Disturbance and man-
agement in larch-cembra pine forests in Europe. In:

Schmidt, Wyman C; McDonald, Kathy J., comps. Pro-
ceedings — symposium on whitebark pine ecoystems:
ecology and management of a high-momitain resoiirce.
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department
of Agricultvire, Forest Service, Intermoxantain Research
Station: 25-36.

Holtmeier, Friedrich-Karl. 1993. European larch in
Middle Europe. In: Proceedings — symposium on ecology
and management of larix forests: a look ahead. Ogden,
UT: U.S. Department of Agriculture, Forest Service, In-
termountain Research Station. [In preparation].

Kleiber, Helga. 1974. Pollenanalytische Untersuchvmgen
zum Eisriickgang und zur Vegetationsgeschichte im
Oberengadin I. Botanisches Jahrbuch fiir Systematik.
94(1): 1-53.

Mattes, Hermann. 1978. Der Tannenhaher im Engadin —

Studien zu seiner okologischen Funktion im Arvenwald.

Miinstersche Geographische Arbeiten. 2. 87 p.
Mattes, Hermann. 1982. Die Lebensgemeinschaft von

Tannenhaher und Arve. Eidgenossische Anstalt fur das

forstliche Versuchswesen. Berichte. 24. 74 p.
Oberdorfer, Erich. 1970. Pflanzensoziologische Exkur-

sionsflora fiir Siiddeutschland. 3d ed. Stuttgart: Verlag

Eugen Ulmer. 987 p.
Schlatter, A. J. 1935. Die Aufforstungen und Verbau-

xmgen des Oberengadins in den Jahren 1875-1934.

Schweizerische Zeitschrift fur Forstwesen. 86: 309-328.

Evolution and
Taxonomy

International Workshop
St. Moritz 1 992

GENETIC DIFFERENTIATION AND
PHYLOGENY OF STONE PINE
SPECIES BASED ON ISOZYME LOCI

Konstantin V. Krutovskii
Dmitri V. Politov
Yuri P. Altukhov

Abstract— F-statistics analysis based on isozyme loci data has
revealed comparatively low levels of intraspecific genetic differen-
tiation among natural populations of Eurasian stone pine species:
Siberian stone pine {Pinus sibirica Du Tour), Korean stone pine
{Pinus koraiensis Siebold et Zucc), and Japanese (dwarf Siberian
or mountain) stone pine {Pinus pumila [Pall.] Kegel). Only about
2 to 4 percent of the total isozyme gene diversity was due to inter-
population variation (FgT = 0.021-0.040), and the overwhelming
part of the total variation, over 96 percent, belonged to intrapopu-
lation variation. These data are discussed in relation to distin-
guishing features of conifer biology, in general, and to a specific
mode of stone pine seed distribution by nutcrackers, in particular.
Cluster analysis of more-studied Siberian stone pine populations
has shown that genetic distances based on isozyme allele frequen-
cies reflect geographical distribution of these populations in spite
of the generally low values of genetic distances (D = 0.001-0.031,
on the average of 0.006).

The authors have studied phylogenetic relationships between all
stone pine species (subsection Cembrae), including one European
stone pine population— Swiss stone pine {Pinus cembra L.) and one
population of the only representative of stone pine species in North
America— whitebark pine {Pinus albicaulis Engelm.), using genetic
distances based on 16 isozjmie loci, whose genetic control has been
resolved for all stone pine species. The results have confirmed
stone pine {Cembrae) species to be a compact group of very closely
related and genetically similar species, having monophyletic ori-
gin, supposedly from ancient Siberian stone pine. The smallest
distance (D = 0.065) has been found between Pinus cembra and
Pinus sibirica that corresponded, mostly, to their subspecies taxo-
nomic status. Other interspecific distances were in good agree-
ment with the species status (D = 0.121-0.268).

As to the controversy on the taxonomy of whitebark pine, this
species undoubtedly belongs to subsection Cembrae (D = 0.121-
0.256). The results of phylogenetic study have been confirmed by
the recently obtained data of cpDNA restriction fragment analyses.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our ICnowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Konstantin V. Krutovskii is Senior Research Scientist, Laboratory of
Population Genetics, N.I. Vavilov Institute of General Genetics, Russian
Academy of Sciences, GSP-1 Moscow 117809 B-333, Russia (current ad-
dress until 1995: Department of Forest Genetics and Tree Improvement,
Georg-August-University, Biisgenweg 2, Gottingen, 37077, Germany);
Dmitri V. Politov is Research Scientist, same laboratory; Yuri P. Altukhov
is Professor, Member of Russian Academy of Sciences, Head of the Labora-
tory of Population Genetics, Director of the Institute of General Genetics.

Five stone pine species are traditionally united in sub-
section Cembrae of section Strobus, subgenus Strobus,
genus Pinus (Critchfield and Little 1966; Shaw 1914).
European or Swiss stone pine {Pinus cembra) is distrib-
uted in the Alps and Carpathians and has the smallest
area among stone pine species. An east Asian stone pine
species, Korean pine (P. koraiensis), is widely distributed
in eastern Russia (Russian "Far East"), Manchuria, and
Japan, but mostly in Korea. Two other stone pine species
are: Japanese stone pine {P. pumila), widely distributed
in northeastern Siberia extending to eastern Asia, as far
as central Japan; and Siberian stone pine (P. sibirica),
widely distributed, mostly in Siberia extending to northern
Mongolia and to the eastern border of Europe. Whitebark
pine (P. albicaulis) is widely distributed in western North
America, extending from British Coliunbia to the southern
Sierra Nevada, and represents the only stone pine species
growing outside Eurasia.

All these species are of great interest for different kinds
of investigations, but some main aspects make these species
very attractive from evolutionary and population genetics
points of view and explain the necessity of their genetic
study.

First, they occupy a vast territory in Eurasia and North
America and most of them belong to the main forest-forming
species. Thus, they have great ecological and economic sig-
nificance, especially in Eurasia. They are subjected to inten-
sive harvesting and, consequently, need special breeding
and gene conservation programs based on genetic data.

Second, stone pine species consist of different kinds of pop-
ulations. For example, Swiss stone pine includes mostly
small, isolated populations and occupies a relatively small
territory. On the other hand, Siberian stone pine includes
huge, continuous populations occupying a very leirge terri-
tory in Siberia. Some of the stone pine species are com-
pletely isolated from others: Swiss stone pine, growing in
Europe, and whitebark pine, growing in North America,
stand apart from Asian stone pine species. At the same
time, Siberian, Korean, and Japanese stone pines have
sympatric zones. Thus, from evolutionary and population
genetics points of view and for gene conservation programs
it would be interesting and important to study genetic dif-
ferentiation, gene dispersal patterns, and mating system
and population genetic structure of these species and com-
pare them with each other.

Third, in spite of the fact that stone pines have a wide
geographical distribution and some morphological and
ecological differences, they are grouped into a separate
subsection Cembrae, because they have some diagnostic

morphological traits that distinguish them from other
Strobus species. Most significant of these are wingless large
seeds and indehiscent cones that remain closed in matu-
rity. These traits are supposed to be the result of long co-
evolution of stone pines and nutcrackers (Nucifraga ssp.),
who play a crucial role in natural reproduction and dis-
tribution of stone pine species (Lanner 1980, 1982, 1990;
Tomback 1983; Tomback and Linhart 1990; Tomback and
others 1990). Nevertheless, taxonomy of these species is
still being discussed. Some specialists consider Swiss stone
pine a subspecies of Siberian stone pine (or vice versa) and
some think that whitebark pine should be included into
subsection Strobi because of its morphological and ecologi-
cal resemblance to some Strobi species (Critchfield 1986;
Millar and Kinloch 1991). Taking all this into account, the
use of genetic markers for study of phylogenetic relation-
ships between stone pine species would be very appropriate
and fruitful.

Thus, the main tasks of our research were estimation of
genetic differentiation and subdivision of stone pine popula-
tions and study of phylogenetic relationships between stone
pine species using isozyme loci as genetic markers.

The advantages and disadvantages of using isozymes in
plant systematics and phylogeny have been fairly thoroughly
discussed in scientific literature (see for example, Crawford
1983; Gottheb 1977; Strauss and others 1992). We would
only like to note here such taxonomically most important
advantages as, first, the possibility of using the same trait
(isozyme electrophoretic pattern encoded by homologous
loci) to classify taxons of almost any phylogenetic level, both
related and unrelated, and second, the possibility of quan-
titative estimation of divergence degree. One of the prob-
lems of traditional systematics is associated with the objec-
tive choice of diagnostic traits, but initial selection of such
traits usually leads to their preliminary "weighting" (evalu-
ation) and inevitably produces subjectivity. It is, therefore,
extremely important to use, above all, such traits in taxo-
nomic studies as can be taken for analysis without "weigh-
ing," more or less arbitrarily, and about whose variability
nothing has been known a priori. Molecular genetic mark-
ers (isozyme loci and DNA markers) are one of the best,
from these points of view.

The study presented here is the first one devoted to study
of phylogeny of the whole subsection Cembrae using isozyme
markers and is only a part of long-term studies on popula-
tion genetics of stone pine species carried out in the Labora-
tory of Population Genetics of N. I. Vavilov Institute of Gen-
eral Genetics (Russian Academy of Sciences, Moscow, Russia)
on the basis of isozyme analysis. More details on inheritance
and linkage of isozjnne loci, mating system, population ge-
netic structure, and other aspects of stone pine population
genetics can be found in our previously published papers
(Krutovskii and Politov 1992; Krutovskii and others 1987,
1988, 1989, 1990; PoHtov 1989; Politov and Krutovskii 1990;
Politov and others 1989, 1992) and in the accompanying pa-
per included in these proceedings by Politov and Krutovskii.

MATERIAL AND SAMPLING SITES

Seventeen single-tree and two bulk seed samples from
natural populations have been used for isozyme analysis in
our study. Eleven localities of Siberian stone pine, three

localities each of Korean and Japanese stone pines, and
one locality each of Swiss stone and whitebark pines have
been studied on 16-20 isoz3niie loci (table 1). More details
about seed samples and characteristics of the sampled
natural populations can be found in our other publications
(Krutovskii and Politov 1992; Krutovskii and others 1988,
1989, 1990; Pohtov 1989; Pohtov and Krutovskii 1993;
Pohtov and others 1992).

ELECTROPHORETIC ANALYSIS

Isozymes of seeds have been studied by electrophoresis
in starch gel. Method of electrophoretic analysis, specimen
preparation, the buffer systems used, genetic interpretation
of zymograms, designations of allozymes, alleles, and loci,
and inheritance of enzyme systems have been described
elsewhere (Krutovskii and others 1987; Politov 1989). Geno-
types of trees have been determined by analyzing segrega-
tion of allozyme alleles among their endosperms, which are
haploid in conifers and genetically identical to maternal
tree gametes. Together with adult tree allele frequencies,
pollen allele frequencies inferred from effective pollen gene
pool (paternal alleles of the analyzed seed embryos) have
been used for estimation of genetic distances. These fre-
quencies of successful pollen give probably more represen-
tative estimation of the studied populations' gene pools due
to predominantly outcrossing pollination among stone pine
species (Pohtov and Krutovskii 1990, 1993).

In total, 536 trees fi-om 17 localities (11-75 trees each) and
two bulk seed samples of five stone pine species have been
analyzed on 16-20 isozyme loci. More than 3,500 (7-20 per
tree) embryos have been analyzed on a smaller amount of
isoz5Tne loci because of poor resolution of some enzyme sys-
tems in embryo tissue. A relatively small number of trees
have been analyzed in some populations and species, but
according to some theoretical (Nei 1987a,b) and experimen-
tal investigations (Shurkhal and others 1992), the number
of loci used in interspecific phylogenetic analysis is more
important than the sample size. We have studied phyloge-
netic relationships between all stone pine species using ge-
netic distances based on 16 isozyme loci, Adh-1, Dia-2, Fe-2,
Gdh, (iot-1, Got-2, Got-3, Idh, Mdh-1, Mdh-2, Mdh-3, Mdh-4,
Pgi-2, Pgm-1, Pgm-2, and Skdh-1, whose allelic variation has
been unequivocally determined for all stone pine species.
Such a number of loci is used commonly and gives quite
reasonable phylogenetic trees (Nei 1987a,b).

DATA ANALYSIS

Determination of parameters of intra- and interpopula-
tional genetic diversity, estimation of genetic differentia-
tion and genetic distances, and clustering and construction
of dendrograms for stone pine populations have been car-
ried out by using IBM PC version 1.7 of the computer pro-
gram BIOSYS-1 (Swofford and Selander 1981) and IBM
PC-program RESTSITE, version 1.1 (Nei and Miller 1990).

Using our data, we have estimated different kinds of ge-
netic distances and tried different methods of clustering.
Methods for calculating genetic distances and methods of
clustering differ in their assumptions concerning homoge-
neity of evolutionary rates or proportionality to the time of
divergence. Consequently, topology of derived phylogenetic

Table 1 — Name, abbreviated designation, sample size, and geographic location of sampling sites

Locality

Designation

Number
of trees

Region

Siberian Stone Pine

Filin Klyuch

Mutnaya Rechka M

Maiyi Kebezh MK

Sobach'ya Rechka SR

Listvyanka L

Yailyu YA

Smokotino S

Zorkartsevo-1 Z-1

Zorkartsevo-3 Z-3

Turukhansk TU

Nyalino NYA

Ust'-Chornaya UCH

Grustnyi GR
Uttyveem UT
Kamenskoye K

Malokhekhtsirskoye MKH
Khor KHO
Sikhote-Alin' SA

Livingstone Falls LF

75
15
12
41
43

43
34
34

MOO

Western Sayan Mountains (Ermakovskii
District, Krasnoyarsk Territory,
eastern Siberia)

Altai Mountains (Gorno-Altaiskaya

Autonomous Region, eastern Siberia)
Tomskaya Region (western Siberia)

'88

North of Krasnoyarsk Territory,
(Turukhanskii District, North of
eastern Siberia)
Khanty-Mansi Autonomous District
(Tyumenskaya Region, North of
western Siberia)

Swiss Stone Pine

15 Eastern Carpathians (western Ukraine)

Japanese Stone Pine

54 North of Kamchatka Peninsula

Korean Stone Pine

30
16
11

Whitebark Pine

Khabarovsk Territory

(Russian "Far East")

Primorskii Territory (Russian "Far East")

Rocky Mountains Forest Reserve,
Bow-Crow Forest (southwestern
Alberta, Canada)

'Number of analyzed seeds from mixed sample obtained from large number (>100) of trees.

trees based on different distances and methods of cluster-
ing can vary significantly for the same data.

Among methods of clustering that assimie homogeneity
of evolutionary rates, an unweighted pair group method
using arithmetic averages, UPGMA, is the most widely
used one (Sneath and Sokal 1973). Among methods that
do not assume homogeneity of evolutionary rates, the dis-
tance Wagner procedure (Farris 1972) and neighbor-joining
method, NJM (Saitou and Nei 1987) are the most widely
used ones.

The most common genetic distances used in isozjnne
studies of trees are Nei's (Nei 1972, 1978), Rogers' (1972),
and CavalU-Sforza and Edwards' (1967) distances. How-
ever, Nei's distance, D, along with advantages has some
disadvantages, the main one being that it is not metric
because it does not obey the triangle inequality and, conse-
quently, cannot be used in the distance Wagner procedure
or NJM. Thus, we have used D (Nei 1972) only in UPGMA
dendrograms.

Taking into accoimt: (1) that homogeneity of evolution-
ary (or allozyme allele substitution) rates can usually be
safely assumed in comparing conspecific populations of
widespread, continuous tree species with high gene flow,
like stone pine populations; (2) shortage of space for pre-
sentation of all obtained dendrograms; and (3) possibility
for wider comparison vnth published data, we will present
only UPGMA dendrograms based on Nei's distance, D,
for analysis of intraspecies stone pine population relation-
ships. However, to avoid problems connected with possible
nonhomogeneity of evolutionary rates of different species,
we will present NJM dendrograms of stone pine species
based on Cavalli-Sforza and Edwards' (1967) chord dis-
tances, Dch, in addition to UPGMA dendrograms based on
D and D.^.

Nei's genetic distances matrix has also been used for
principal coordinate analysis of Siberian stone pine popula-
tions using the computer program NTSYS-pc (Rohlf 1988),

INTRASPECIFIC DIFFERENTIATION

Levels of Genetic Differentiation Between
Populations — Distribution of intraspecific genetic variation
between populations of stone pines has been analyzed for
three species studied in several localities using F-statistics
(Nei 1977). Levels of genetic differentiation among popula-
tions were comparatively low — only about 2-4 percent of
the total intraspecific isozyme gene variation was due to
interpopulation variation (FgT = 0.021-0.040), and the over-
whelming part of the total variation, over 96 percent, be-
longed to intrapopulation variation (table 2). Loci Fe-2,
Mdh-3, and Skdh-1 of Siberian stone pine, Adh-1, Adh-2,
Gdh, Got-2, Mdh-2, Mdh-3, Mdh-4, and Skdh-2 of Japanese
stone pine, and Mdh-4 of Korean stone pine make a most
significant contribution in intraspecific differentiation of
these species (table 2). The geographic variation of these
loci may be adaptive and should be studied in detail in fu-
ture investigations.

Nei's genetic distances between populations were also
small, D = 0.001-0.031 and 0.012, on the average (table 3).
These values correspond perfectly to the calculations car-
ried out using references on 22 pine species (Altukhov and
others 1989). The genetic distances between closely located
(<20 km) populations of seven pine species equaled 0.009,
on the average, and between widely dispersed populations
of 15 pine species reviewed — D = 0.036, on average.

Low levels of interpopulation differentiation of allozyme
loci are usual for conifers. Such typical factors for most coni-
fer populations as outcrossing, wind pollination, seed dis-
persal by wind (most conifers) or by birds (stone pines, some
white pines, and pinyons), wide continuous ranges, high
population density, and effective size are considered to re-
duce the influence of genetic drift and, therefore, decrease

heterogeneity of allele frequencies and interpopulation
genetic differentiation (Hamrick 1983; Hamrick and Godt
1989; Hamrick and Loveless 1986; Hamrick and others
1981; Loveless and Hamrick 1984).

We have also simmiarized data on interpopulation genetic
differentiation of 28 conifer species of eight genera estimated
using FgT or analogous GgT parameters of differentiation
(Politov and others 1992). In spite of a, generally, compar-
atively low differentiation among populations, we have
found a quite pronounced tendency to a significantly smaller
differentiation (Fst = 0.02-0.08) among widespread species
with continuous population ranges than among species with
interrupted ranges and small, and often isolated, popula-
tions (Fg-r = 0.11-0.13). Moreover, according to our review,
samples from the same population (samples fi'om localities
within several kilometers radius) revealed, commonly, an
even much lower level of differentiation (Fg^ = 0.006-0.02).

The three stone pine species studied belong to the first
group of species with mainly continuous population range.
Besides, four of the 11 Siberian stone pine samples, F, M,
MK, and SR, have been collected within 20 km radius from
obviously the same population of Malyi Kebezh River Basin
(eastern Siberia, table 1); and the other two samples have
been taken fi^om two stands near Zorkal'tsevo village (west-
em Siberia), Z-1 and Z-3, which are separated from each
other by less than 3 km. The other localities are rather re-
mote from one another, but all of them represent only the
central, continuous part of the species area. There are no
significant barriers to gene flow in this area.

The low differentiation observed in our study may also
be due to the limited number of samples taken, mostly,
from more or less optimal zones of the stone pines area,
impeding the diversifying selection action and promoting
the balancing one (Altukhov 1990).

Table 2 — The estimates of F-statistics for populations of three stone pine species and for all five species

Locus

Siberian stone pine

Korean stone pine

Japanese stone pine

All

Fst

F,s

Fst

Adh-1

-0.030

-0.006

0.023

0.078

0.120

0.045

0.306

0.321 '

0.022

'0.563

Adh-2

{')

(')

{')

-.068

-.046

\020

Dia-1

-.032

-.007

.024

(')

{')

Dia-2

-.060

-.053

.006

{')

(')

-.038

-.013

\025

'.028

Fe-2

-.061

-.012

\046

-.067

-.021

.043

-.105

-.097

.007

'.190

Gdh

-.012

-.001

.010

-.098

.000

.084

.140

^061

'.637

Got-1

-.013

-.001

.012

{')

.002

Got-2

-.012

-.001

.010

{')

-.050

-.022

\027

'.016

Got-3

(')

{')

.238

.241

.004

.014

.015

.001

'.596

Lap-2

.015

.025

.009

-.071

-.023

.045

-.048

-.047

.001

Lap-3

.066

.079

.013

(')

{')

-.064

-.050

.013

Mdh-2

.214

.231

.023

-.251

-.225

.020

-.004

.015

^019

'.240

Mdh-3

-.049

.000

\047

-.067

-.021

.043

-.051

.084

M28

'.118

Mdh-4

-.068

-.055

.012

.017

.104

'.089

-.021

-.013

\ooa

'.362

Pgi-2

.114

.123

.010

{')

-.111

-.107

.003

'.407

Pgm-1

-.074

.051

.022

-.193

-.160

.028

.177

.189

.016

'.187

Pgm-2

{')

(')

{')

-.014

-.013

.002

'.046

Skdh-1

-.146

-.118

\025

-.316

-.194

.092

-.053

-.051

.002

'.187

Skdh-2

{')

.099

.163

'.071

Mean

-.043

-.018

.025

-.077

-.033

.040

-.002

.020

.021

.396

'This locus has statistically significant heterogeneity of allele frequencies among populations.

n"he genetic base of isozyme variation of this locus is not determined or allelic variation cannot be compared with other species,
^his locus is monomorphic in this species.

Table 3 — Genetic distances, D, (Nei 1972) for five stone pine species

Sibarian

Ivor oan

Species

stone pine

Stone pine

Siberian stone pine

0.006
(0.002-0.016)

Swii^^ stonp oinp

0 065
(0.055-0.077)

Jaoanese stone oine

0.202
(0.176-0.253)

0.233
(0.215-0.265)

0 010
(0.003-0.014)

Korean stone pine

0.235

0.268

0.105

0.009

(0.207-0.261)

(0.264-0.275)

(0.096-0.119)

(0.004-0.014)

Whitebark pine

0.121

0.130

0.185

0.256

(0.111-0.129)

(0.166-0.220)

(0.253-0.261)

'N = number of localities studied.

And finally, the low differentiation between stone pine
populations can be considered the result of a specific mode
of seed dispersal by nutcrackers able to carry their seeds
over very long distances (up to 20 km, according to some
observations). The efficiency of this mode and its influence
on the population structure were repeatedly emphasized
by researchers studying relationships between nutcrackers
and whitebark pine, Swiss stone pine, limber pine {Pinus
flexilis James), and pinyons (subsection Cembroides)
(Carsey and Tomback 1992; Furnier and others 1987;
Lanner 1980, 1990; Schuster and Mitton 1991; Tomback
and others 1990, 1992).

Cluster and Principal Coordinates Analysis —

Despite the low values of genetic distances, UPGMA cluster-
ing of stone pine samples based on these distances reflects,
in general, their geographical origin. More detailed analy-
ses have been performed on Siberian stone pine, which has
been studied more than other stone pines. We can tenta-
tively divide all 11 Siberian stone pine localities into three
regions, according to their ecology and geographic location.
The first of them can be called "Southern Siberia Moimtains"
and includes one population from the Altai Mountains (Ya)
and five localities in the western Sayan (FK, M, MK, SR,
and L); the second can be called "Western Siberia Plain"
and includes populations from Tomsk region (S, Z-1, and
Z-3); the third, "Northern Siberia," includes populations
from Khanty-Mansi Autonomous District (NYa) and from
the north of Krasnoyarsk Territory (TU). The dendrogram
obtained corresponds exactly to this classification (fig. 1).
These three regions form three different genetic groups
(clusters) of populations.

Korean pine samples also form clusters in accordance
with their population localization (two geographically more
closely spaced stands, MKh and Kho, have turned out to
be genetically more similar). The same is true of Japanese
stone pine populations (fig. 1). Principal coordinates analy-
ses based on the matrix of Nei's genetic distances between
Siberian stone pine populations also have been performed.
This multidimensional analysis has produced almost the
same result as the cluster analysis. Genetic relationships
between Siberian stone pine populations, graphically pre-
sented in two-dimensional space of the first two principal
coordinates, have perfectly reflected the real geographical

distribution of these populations (fig. 2). Principal coordi-
nates and cluster analyses complement each other and have
produced almost the same result in our case, but the first,
probably, gives more information about population genetic
relationships due to better data presentation. Multivariate
analysis techniques are supposed to provide, basically, bet-
ter possibility for revealing genetic differentiation in coni-
fers, as compared to univariate methods (Krutovskii and
Politov 1992; Yeh and others 1985).

Statistically significant positive correlation between geo-
graphical and genetic distances has been found for Siberian
stone pine populations (r = 0.844, P < 0.001). Such correla-
tion was, however, absent for samples from closely spaced
localities of the Malyi Kebezh River Basin. The limited
number of samples studied has not allowed such analysis
for other stone pine species, but analogous results have

0.010

0.005
I

L
M
MK

Southern

Siberia

Mountains

SR _

Z-1

Z-3

Western

Siberia

Plain

NYA _

Northern
Siberia

MKH~
KHO_

1 Khabarovsk"
Territory

Primorski! -
Territory

GR
UT

K -

Kamchatka
Peninsula

Siberian

stone pine

Korean
stone pine

Japanese
stone pine

Figure 1 — UPGMA phenograms of stone pine
populations based on Nei's genetic distances,
D, (Nei 1972). (See full names and origins of
populations in table 1 .)

NYA

Southern
Siberia
Mountains

Western
Siberia Plain

-0.15

Figure 2 — Projections of Siberian stone pine
populations in two-dimensional space of the first
two principal coordinates, X and Y, according to
Nei's genetic distances, D, (Nei 1972). (See full
names and origins of populations in table 1 .)

also been obtained for some other pine species (Furnier
and Adams 1986; Steinhoff and others 1983; Wheeler and
Guries 1982), including, probably, that isolation by distance
(Nei 1975), introgressive hybridization (Millar 1983), or
gradiently changing selection (Bergmann 1978) could play
a significant role in genetic differentiation of some pine
species.

We have also analyzed correlation between geographical
and Nei's distances calculated for Siberian stone pine popu-
lations using data on individual loci (allele frequencies of
eight loci in the effective pollen pool have been used). For
loci Adh-1, Dia-2, Fe-2, Pgm-1, and Skdh-1 correlation was
significant. The other loci (Lap-3, Mdh-2, and Pgi-2) dis-
played lack of correlation. This may suggest a selective role
of certain loci. Patterns of spatial differentiations seem to
be determined by complex interaction of gene flow and selec-
tion. Distribution of genetic variability may be also affected
by "evolutionary footprints," such as, for example, Siberian
population descent from different refugiums.

INTERSPECIFIC DIFFERENTIATION

Genetic Distances Between Stone Pine Species —

We have estimated both genetic distances, D^j, and D, be-
tween all stone pine species based on 16 isozyme loci data,
but only D values are presented in table 3, because of space
limits. The smallest genetic distances have been found
between Siberian and Swiss stone pines (D = 0.052-0.077).
They are just a little larger than those between subspecies
of lodgepole pine (Pinus contorta Dougl. ex Loud.), D = 0.008-
0.019, approximately 0.012 on the average (Wheeler and
others 1983), similar to the distances between subspecies
of brutian pine {Pinus brutia Ten.), D = 0.03-0.09 (Conkle
and others 1988), and a little smaller than distances be-
tween two very closely related species, jack pine (Pinus
banksiana Lamb.) and lodgepole pine, which have a wide
zone of introgressive hybridization, D = 0.097, according to

the study by Dancik and Yeh (1983), or D = 0.108, according
to Wheeler and Guries (1987). Nei's distances estimated
between other pairs of stone pine species (D = 0.121-0.268,
table 3) are close to the values reported for pine species
of subsections Contortae: 0.078-0.158 (Wheeler and others
1983) and Oocarpae: 0.11-0.27 (Millar and others 1988).
Approximately the same value, D = 0.27-0.43, has been
reported for Pinus brutia-Pinus halepensis comparison
(Conkle and others 1988). We have estimated the average
genetic distance between species fi-om the same subsections
using a few references (Conkle and others 1988; Millar and
others 1988; and some others). It equaled 0.185, on the av-
erage, for 14 pine species and surprisingly corresponded to
the average distance estimated for all stone pine species in
our study — D = 0.180, and for four Eurasian stone pine spe-
cies studied by Shurkhal and others (1991a,b)— D = 0.138
(for comparison, D = 0.185, on the average, for four Eurasian
stone pines used in our study). The small discrepancy be-
tween the latter two estimates is not surprising, taking into
account the sampling strategy used by Shurkhal and others
(1991a,b; 1992). They have analyzed only a few trees of
each species (mostly, from Botanical Garden collections),
which should inevitably bias genetic distance estimates to
smaller values, due to a higher probability of random fixa-
tion of most common, frequent alleles in small samples of
trees from different species.

Levels of interspecific genetic differentiation obtained
between stone pines strongly confirm their taxonomical
status as closely related species of the same subsection.
Nei's genetic distances obtained for species belonging to
different subsections or sections are usually higher than
0.3 and vary between 0.3 and 0.8 (Shurkhal and others
1991a,b; 1992). According to phylogenetic analysis carried
out by Shiu-khal and others (1991a,b; 1992) using isoz)Tne
loci, fotir stone pine species of subsection Cembrae (white-
bark pine has not been analyzed) and three white pine spe-
cies of subsection Strobi, Balkan or Macedonian pine (Pinus
peuce Griseb.), blue or Himalayan pine (Pinus griffithii
McCleland), and eastern white or We3maouth pine (Pinus
strobus L.), form separate clusters. The Nei's genetic dis-
tance between these two very closely related subsections
equals 0.278, on the average (calculated by us, using the
data in table 5 from Shurkhal and others 1991b). This is
almost two times larger than that between Eurasian stone
pines (D = 0.138).

The lack of species-specific diagnostic loci among stone
pines also confirms the correctness of uniting them in the
same separate subsection Cembrae. Only Siberian and
Korean stone pines, on locus Gdh, and Swiss and Korean
stone pines, on locus Mdh-4, do not have common alleles.
Besides, Korean stone pines have a unique electrophoretic
pattern of LAP enzyme indicating the existence (or expres-
sion) of an additional Lap-4 locus.

Loci Adh-1, Gdh, Got-3, Mdh-4, and Pgi-2 make a most
significant contribution to interspecific differentiation of
stone pines, but totally only 39 percent of the whole sub-
section Cembrae genetic variation belongs to interspecific
differences (table 2).

Thus, our data give a strong genetic evidence of common,
apparently monophyletic stone pines origin and compara-
tively recent time of their divergence. Subsection Cembrae
seems to be a valid taxon.

Phylogenetic Relationships Inferred From Genetic
Distances — Many publications have been devoted to stud-
ies of taxonomic and phylogenetic relationships between
stone pines using morphological, physiological, ecological,
and other traits. Our study was the first one to use genetic
markers, isozyme loci, for analysis of phylogenetic relation-
ships between all five stone pine species. These species are
traditionally united in subsection Cembrae, according to
some characteristic features (Critchfield and Little 1966;
Farjon 1984; Little and Critchfield 1969; Shaw 1914). The
overwhelming majority of researchers consider the main spe-
cific features of this subsection, indehiscent cones and wing-
less seeds, to be inherited fi*om a common ancestor. These
traits seem to be a relatively recent acquisition. Their ap-
pearance is considered to be a result of coevolution of stone
pines and nutcrackers of genus Nucifraga foraging their
seeds and playing the main role in their maintenance and
dispersal (Lanner 1980, 1982, 1990; Tomback 1983;
Tomback and Linhart 1990; Tomback and others 1990).

Thus, the subsection Cembrae is regarded as a natural
monophyletic taxon. Nevertheless, there also exists a dif-
ferent point of view (Critchfield 1986), according to which
traits shared by all stone pines are insufficiently confirmed
as homologous. Data on artificial hybridization appear to
be rather contradictory (Critchfield 1986). The subsection
is believed to have polyphyletic origin and, therefore, is
not a valid taxon (see for discussion, Lanner 1990). Millar
and Kinloch (1991) also have pointed out a nimiber of im-
solved problems in taxonomy and phylogeny of subsection
Cembrae.

Molecular and biochemical genetic markers have proven
to be a useful tool for study of biosystematic and phyloge-
netic relationships in conifers (Strauss and others 1992).
We believe that application of biochemical genetic markers
like isozymes can help shed light on the evolution of stone
pine species. The dendrograms produced by NJM and
UPGMA clustering using D^h and D matrix (table 3) based
on 16 isozyme loci have apparently revealed real phyloge-
netic relationships between five stone pine species (fig. 3).

According to this dendrogram, Siberian and Swiss stone
pines appear to be a most closely related pair of species
(D = 0.065). This was not imexpected, because of the well-
known surprisingly high morphological similarity (Bobrov
1978) and successful artificial hybridization (Critchfield
1986) between them. Practically all traditional botanical
classifications note the absence of diagnostic morphological
traits, which means that interspecific differences are more
quantitative than qualitative. According to paleobotanical
(mainly palynological) data, Bobrov (1978) considered these
two species as two now separately located marginal parts
of a former single ancestral species spread on a vast range
area, from the western Alps to the northeastern Yakutia
(eastern Siberia), whose pollen has been foimd in Paleocene,
Pleistocene, and later deposits.

As Bobrov (1978) indicates, "Siberian stone pine must be
regarded as the initial, more ancient species, formerly more
widely distributed, whereas the West European stone pine,
Pinus cembra, is evolutionarily still very yoimg and has be-
come geographically isolated in the recent past, being, there-
fore, hardly disting\iishable from its Siberian ancestors."

Breakup of this ancient range, its division into two parts,
and decrease of its western part occurred, most likely, as

UPGMA

Isozyme loci

-L

225 200 175 150 125 100 75 50 25

UPGMA

cpDNA RFLP

Whitebark pine
Swiss stone pine
Siberian stone pine
Korean stone pine
Japanese stone pine

D (X10-3)

Whitebark pine
Swiss stone pine
Siberian stone pine
Korean stone pine
Japanese stone pine

90 80 70 60 50 40 30 20 10

UPGMA

isozyme loci

d (X10-5)

Whitebark pine
Swiss stone pine
Siberian stone pine
Korean stone pine
Japanese stone pine

405 360 315 270 225 180 135 90 45

Dch(x10-3)

NJM isozyme loci

Siberian stone pine

C Siberian

Swiss stone pine

Whitebari< pine

— — Korean stone pine
-■i-v— i— i^^— Japanese stone pine

-J \ \ lD,,(x10-3)

45 90 135 180 225 270 315 360 405

NJM

Swiss stone pine
Siberian stone pine

Korean stone pine

cpDNA RFLP

Whitebark pine

I L

J L

Japanese stone pine
-Jd (X10-5)

Figure 3 — UPGMA and NJM dendrograms of five
stone pine species (subsection Cembrae) based on
Cavalli-Sforza and Edwards' (1967) and Nei's (Nei
1972) genetic distances, and D, calculated from
isozyme data, and on d (Nei 1 987b; Nei and Miller
1990), calculated from chloroplast (cp) DNA restric-
tion fragment length polymorphism (RFLP) data
(Krutovskii and Wagner 1993).

late as in the Holocene and was associated with its substi-
tution in the central part of the area by other species due
to the global warming. The time of divergence (10,000 to
25,000 years), estimated from the genetic distance between
Siberian and Swiss stone pines corresponds, approximately,
to these events (for more details see Krutovskii and others
1990). Thus, the low level of divergence between Siberian
and Swiss stone pines can be explained by the short time
of isolation.

Nevertheless, a considerable amount of genetic difference
has accumulated during this time — genetic distances be-
tween Siberian and Swiss stone pines are approximately
10 times larger than those between populations of the same
species (table 3) and are comparable with distances reported
by Conkle and others (1988) for subspecies and very closely
related pine species. Our data have also been recently con-
firmed by Goncharenko and others (1992) and Shurkhal and
others (1991a,b; 1992). Their estimates of Nei's genetic dis-
tance between Siberian and Swiss stone pines equaled
0.030 and 0.042, correspondingly. Thus, Swiss stone pine
seems to be a highly diverged geographical race or subspe-
cies of Siberian stone pine, or a species in statu nascendi.

Unexpectedly relatively small differentiation found be-
tween Japanese and Korean stone pines (D = 0.105) makes
us search for a new approach to understanding their origin.
These species, in fact, have little in common in morphology
of both vegetative and generative organs, but according to
the study of parenchjnnal tissue of needle cell walls, stone
pine species and some closely related species can be divided
into two t3T)es: Siberian and Swiss stone pine form the first
type, while Japanese and Korean stone pines belong to the
second type, but to different subtypes (Litvintseva 1974).
Perhaps, the evolutionary pathways of these species have
truly separated from a common ancestor relatively recently.
Another explanation of their genetic similarity is gene ex-
change through introgressive hybridization, which could
have occurred (or seldom occurs), taking into account the
present sympatry of these species and, as a rule, poorly
developed mechanisms of reproduction isolation between
closely related pine species. More detailed discussion can
be foimd in Krutovskii and others (1990), but recently ob-
tained data also confirm the division of Eurasian stone pines
by isozyme loci analysis into two groups: Siberian and Swiss
stone pines (D = 0.030-0.042), on one hand, and Korean and
Japanese (D = 0.089-0.143), on the other hand (Goncharenko
and others 1992; Shurkhal and others 1991a,b; 1992).

A most interesting problem, from our point of view, is ex-
planation of the whitebark pine's closer affinity to Siberian
and Swiss stone pines (D = 0.121-0.130) than to Japanese
stone pine (D = 0.185), although quite opposite relation-
ships could be suspected considering these species' present
area. Among stone pines Siberian and Swiss stone pines
are, probably, genetically most similar to the original an-
cient stone pine, due to large refugiums in the Alps and
Siberian (Altai and Sayan) Mountains, which could help
them save a great deal of ancient genetic variation; and
whitebark pine also could keep more genetic similarity with
the ancient species than Korean and Japanese stone pines.
Another explanation could be that stabilizing selection sup-
ports genetic similarity between whitebark pine and Siberian
and Swiss stone pines, because whitebark pine ecologically

has, probably, more in common with them than with Korean
and Japanese pines.

However, the dendrogram obtained by NJM, and based
on Cavalli-Sforza and Edwards' (1967) distance, D^^, has
revealed a kind of "intermediate" position of whitebark pine
between two stone pine groups, Swiss and Siberian stone
pines, on one hand, and Korean and Japanese stone pines,
on the other (fig. 3). In these cases NJM gives more an in-
tuitively realistic phylogenetic tree. It is important to note
that both UPGMA dendrograms based on D^,, and Nei's dis-
tance, D, have practically the same topology, but they differ
from the NJM dendrogram. Thus, one can conclude that
assumption of equal evolutionary rates of stone pine spe-
cies can be erroneous.

Millar and Kinloch (1991) discuss several theoretically
possible variants of whitebark pine origin: (1) from ancestral
Eurasian stone pine with penetration into North America
during the Mesozoic (before Laurasia segregation). Ter-
tiary, or Quartemary (through the Bering land bridge) time;
(2) fix)m one of the American pines of subsection Strobi. Now
we can finally solve this problem. Nei's genetic distances
(D) between whitebark pine and Eurasian stone pines equal
approximately 0.121-0.256 (table 3). The time of divergence
it) between whitebark pine and Eurasian stone pines, theo-
retically estimated from these values, is 0.6-1.3 million years,
using calibration coefficient (k) 5x10^ in the equation t = kD
(Nei 1975; 1987a,b). This corresponds quite well to the ap-
proximate geological time of the Bering Strait opening —
1.8-3.5 million years ago — estimated by different methods
(see references in Grant 1987). Consequently, this event,
the Bering Strait opening, can be, obviously, considered
as the beginning of genetic differentiation between North
American and Eurasian pines due to geographical isolation.

Thus, the hypothesis of a relatively recent penetration of
ancient stone pines into North America through the Bering
lemd bridge in the Pliocene seems most preferable. If we
assumed the hypothesis that a more ancient penetration
of stone pines through a "Euramerican" connection had
occurred during Mesozoic time (Millar and Kinloch 1991),
much more differentiation between whitebark pine and
Eurasian stone pines should have been found in our study.
The assumption of whitebark pine origin from Strobi pines
must be rejected, according to our data, but phylogenetic
relationships of stone pines with other relative species from
subsections Strobi and Cembroides certainly need much
more analysis. The first steps in this direction have been
made recently using cpDNA markers.

Chloroplast DNA Data and Pine Phylogeny— Only
a few studies on pine phylogeny using DNA markers have
been made. This allows us to draw a short comparison of
the Millar and Kinloch (1991) results with our data, pre-
sented here, paying special attention to phylogenetic rela-
tionships of stone pines with some other closely related
species.

Chloroplast DNA (cpDNA) restriction fragment length
polymorphism (RFLP) of the five stone pine species and four
white pine species of a closely related subsection Strobi —
limber pine, sugar pine (P. lambertiana Dougl.), western
white pine (P. monticola Dougl.), and eastern white pine —
have been recently studied by Krutovskii and Wagner (1993).
Data have been obtained by molecular hybridization of 27

probes, representing the entire chloroplast genome of
P. contorta, with restriction fragments from 20 endonu-
cleases. The total nimiber of restriction fragments per
sample varied from 378 to 422.

Chloroplast DNA variation among trees within species
was not pronounced (D = 0.0-0.0013) and was comparable
with variation between populations within species, but
cpDNA diversity, in general, was quite high — 50 to 100
percent individual cpDNA belonged to different haplotypes
(for more details see Krutovskii and Wagner 1993). The
dendrogram constructed on Nei's genetic distances (D),
estimated from these data (Nei 1987b; Nei and Miller
1990), has shown almost the same phylogenetic relation-
ships between stone pine species as established on isozyme
loci data (fig. 3). The stone and white pine species have
been imdoubtedly divided into two separate groups (fig. 4).
It is interesting to note that NJM dendrograms give better
correspondence between cpDNA and isozyme data. Com-
paring NJM and UPGMA dendrograms we can suppose
that assumption of homogeneity of evolutionary rates is
less proper for cpDNA than for nuclear (isozyme) genes.

The cpDNA data also indicate that genetic differentia-
tion is more developed among the four Strobi species than

among the Cembrae species. The same concliision has been
made by Shurkhal and others 1991a,b; 1992) on the isozyme
data.

Chloroplast DNA's of whitebark pine and Eurasian stone
pine species were remarkably similar. This similarity, to-
gether with clear cpDNA divergence between the two sub-
sections, confirms systematic classification of whitebark
pine in subsection Cembrae, and not in subsection Strobi.
Phylogenetic relationships of whitebark pine with Siberian
and Swiss stone pines revealed by cpDNA RFLP analysis
are surprisingly consistent with the conclusions from the
analysis of 16 isozyme loci (fig. 3). Thus, ovir observation
is not an artifact, because two independent sets of genetic
markers (isozymes and cpDNA RFLPs) have produced the
same conclusions. Moreover, Szmidt (1991) notes that the
present geographic distribution of some forest tree species
often shows little correlation with the degree of cpDNA dif-
ferentiation among them, because the present geographic
distribution of some species is of relatively recent origin.

The data obtained by Szmidt and others (1988), who used
four restriction enzymes for cpDNA RFLP analysis of 12
pine species, including two stone pine species (Swiss stone
and whitebark pines) and two white pine species (western

UPGMA

cpDNA RFLP

Whitebark pine
Swiss stone pine
Siberian stone pine
Korean stone pine
Japanese stone pine
Western white pine
Limber pine
Sugar pine
Eastern white pine

d (xlO-3)

Stone
Pines

White
Pines

NJM

J Sit

Siberian stone pine
Swiss stone pine
Whitebark pine
Korean stone pine
apanese stone pine

Western white pine
Limber pine
Sugar pine

Stone
Pines

Eastern white pine
I id(x10-3)

White
Pines

Figure 4 — UPGMA and NJM dendrograms of five stone pine spe-
cies (subsection Cembrae) and four white pine species closely re-
lated to them: Pious flexilis, P. lambertiana, P. monticola, and P.
strobus (subsection Strobi), based on Nei's genetic distances, d,
(Nei 1987b; Nei and Miller 1990), and calculated from chloroplast
(cp) DNA restriction fragment length polymorphism (RFLP) data
(Krutovskii and Wagner 1993).

and eastern white pines), also confirm our results. They
evaluated phylogenetic relationships between these species
using UPGMA clustering based on dissimilarity index, anal-
ogous to Nei's D. According to the dendrogram given by
Szmidt £ind others (1988), Swiss stone and whitebgirk pines
form a very tight separate cluster. Western white pine is
distantly linked to this cluster, which exactly corresponds
to our data (fig. 4). Eastern white pine forms a separate
branch but is closely linked to the group of Swiss stone,
whitebark, and western white pines.

Analogous data have been obtained by Strauss and
Doerksen (1990) who studied one sample fi"om each of 18
to 19 pine species representing 14 to 15 subsections using
eight restriction enzjnnes and 17 cloned DNA fragments
(mostly fi"om Pseudotsuga menziesii). At least 65 percent
of the chloroplast genome was covered by the combination
of the probes used. The samples have also included two
stone pine species (Korean stone and whitebark pines) and
two white pine species (blue and sugar pines). Distances
between the samples were estimated via shared characters.
The stone £uid white pine species were extremely similar
to one another, with distances ranging fi^om 0.0 to 0.018.
They form one cluster. The authors have mentioned (data
have not been presented) that other, more limited, restric-
tion fi'agment studies of section Strobus, including three
stone (Swiss, Siberian, and Japanese) and three white pines
(Armand's [P. armandii Franch.], limber, and western white),
have given the same indication of high uniformity. Unfor-
tunately, Strauss and Doerksen (1990) have not been able
to distinguish stone and white pines, probably because a
limited number of restriction enzymes or DNA probes have
been used in their research.

Certainly, stone pines and other closely related species
need more genetic analysis using both isozyme and DNA
markers. Many fascinating aspects of their evolution are
sure to remain hidden, and many discoveries are still w£dt-
ing for research.

CONCLUSIONS

The main conclusions based on our data are:

1. Both biochemical (isozymes) and molecular (cpDNA)
genetic markers are very useful and informative for study
of population genetic structure and phylogeny of stone pine
and other closely related species, and they complement each
other.

2. In spite of the high level of isozyme loci and cpDNA
diversity among stone pine species, genetic differentiation
between the populations is not pronounced, but specific.
As a minimum, the genetic differentiation reflects their
geographical and ecological peculiarities. Thus, it is neces-
sary to study populations fi'om the whole of the stone pine
species area for careful description of gene distribution,
taking into account even very small genetic differences
between populations for developing gene conservation and
reforestation programs.

3. According to both isozyme and cpDNA restriction fi'ag-
ment analyses, P. albicaulis undoubtedly belongs to Cembrae
species and not to Strobi species.

4. Stone pine species consist of a compact group of very
closely related and genetically similar species, obviously

having monophyletic origin, supposedly, fi*om ancient
Siberian stone pine. Thus, subsection Cembrae should
be considered as a valid taxon.

ACKNOWLEDGMENTS

We are very gratefiil to V. N. Vorobjev and N. A. Vorobjeva
(Institute of Ecology of Natural Complexes, Tomsk, Russia),
L. I. Milyutin and his laboratory staff (Institute of Forest
and Timber, Krasnoyarsk, Russia), G. R. Furnier (Univer-
sity of Minnesota, Minneapolis, MN, U.S.A.), R, T. Gut (For-
est Technology Institute, Lvov, Ukraine), and P. Razumov
(Far East Forest Institute, Khabarovsk, Russia) for their
help in collecting and kindly providing Cembrae species seed
samples for isozjnne analysis. We are th£inkful to Yu. S.
Belokon' (Institute of General Genetics, Moscow, Russia)
for very helpfiil assistance in electrophoretic ansdysis, R. M.
Lanner (Utah State University, Logan, UT, U.S.A.), D. F.
Tomback (University of Colorado at Denver, Denver, CO,
U.S.A.), and C. Millar (Institute of Forest Grenetics, Berkeley,
CA, U.S.A.) for very useful comments on pine phylogeny.
We also appreciate H.-R. Gregorius' invaluable comments
that helped to substantially improve the manuscript.

REFERENCES

Altukhov, Y. P. 1990. Population genetics: diversity and sta-
bility. New York: Harwood Academic Publishers. 352 p.

Altukhov, Y. P.; Krutovskii, K V.; Dukharev, V. A.;
Larionova, A. Y.; Politov, D. V.; Ryabokon', S. M. 1989.
Biochemicgd population genetics of woody forest species.
In: Iroshnikov, A. I., ed. International symposium on for-
est genetics, breeding and physiology of woody plants:
proceedings; 1989 September 24-30; Voronezh. Moscow:
The State Forest Committee of the USSR: 21-29.

Bergmginn, F. 1978. The allelic distribution of an acid phos-
phatase locus in Norway spruce {Picea abies) £dong simi-
lar climatic gradients. Theoretical and Applied Genetics.
24: 57-64.

Bobrov, E. G. 1978. Forest-forming conifers of the USSR.
Moscow: Nauka. 190 p. [In Russian].

Carsey, K. S.; Tomback, D. F. 1992. Growth form distribu-
tion and genetic relationships in tree clusters of Pinus
flexilis, a bird-dispersed pine. Oecologia. [In press.]

Cavalli-Sforza, L. L.; Edwards, A. W. F. 1967. Phylogenetic
analysis: models and estimation procedures. American
Journal of Human Genetics. 19(2): 233-257.

Conkle, M. T.; Schiller, G.; Grunwald, C. 1988. Electro-
phoretic analysis of diversity and phylogeny of Pinus
brutia £ind closely related taxa. Systematic Botany.
13(3): 411-424.

Crawford, D. J. 1983. Phylogenetic and systematic infer-
ences fi'om electrophoretic studies. In: Tanksley, S. D.;
Orton, T. J., eds. Isoz)nnes in plant genetics and breed-
ing. Part A. Amsterdam: Elsevier Scientific Publishers:
257-287.

Critchfield, W. B. 1986. Hybridization and classification
of the white pines (Pinus section Strobus). Taxon. 35(4):
647-656.

Critchfield, W. B.; Little, E. L. 1966. Geographic distribu-
tion of the pines of the world. Washington, DC: U.S. De-
partment of Agriculture. 99 p.

Dancik, B. P.; Yeh, F. C. 1983. Allozyme variability and
evolution of lodgepole pine {Pinus contorta var. latifolia)
and jack pine (Pinus banksiana) in Alberta. Canadian
Journal of Genetics and Cytology. 25: 57-64.

Farjon, A. 1984. Pines: drawings and descriptions of the
genus Pinus. Leiden: E. J. Brill. 204 p.

Farris, J. S. 1972. Estimating phylogenetic trees from dis-
tance matrices. American Naturalist. 106(951): 645-668.

Fumier, G. R.; Adams, W. T. 1986. Geographic patterns of
allozyme variation in Jeffrey pine. American Journal of
Botany. 73(7): 1009-1015.

Fumier, G. R.; Knowles, P.; Clyde, M. A.; Dancik, B. P. 1987.
Effect of avian seed dispersal on the genetic structure of
whitebark pine populations. Evolution. 41(3): 607-612.

Goncharenko, G. G.; Padutov, V. E.; Silin, A. E. 1992. Popu-
lation structure, gene diversity, and differentiation in natu-
ral populations of cedar pines (Pinus subsection Cembrae,
Pinaceae) in the USSR. Plant Systematics and Evolu-
tion. 182(3-4): 121-134.

(rottlieb, L. D. 1977. Electrophoretic evidence and plant
systematics. Annals of Missouri Botanical Garden. 64:
161-180.

Grant, W. S. 1987. Genetic divergence between congeneric
Atlantic and Pacific Ocean fishes. In: Ryman, N.; Utter,
F., eds. Population genetics and fishery management.
Seattle, WA: University of Washington Press: 225-246.

Hamrick, J. L. 1983. The distribution of genetic variation
within and among natural plant populations, hi: Chambers,
S.; Schonewald-Cox, C, eds. Genetics and wild population
management. Reading, PA: Addison- Wesley: 335-349;
501-509.

Hamrick, J. L.; Godt, M. J. 1989. Allozyme diversity in plant
species. In: Brown, A. H. D.; Clegg, M. T.; Kahler, A. L.;
Weir, B. S., eds. Plant population genetics breeding and
genetic resources. Sanderland, MA: Sinauer Associates:
43-63.

Hamrick, J. L.; Loveless, M. D. 1986. The influence of seed
dispersal mechanisms on the genetic structure of plant
populations. In: Estrada, J.; Fleming, T. H., eds. Frugi-
vores and seed dispersal. Dordrecht, The Netherlands:
Dr. W. Junk Publishers: 211-223.

Hamrick, J. L.; Mitton, J. B.; Linhart, G. B. 1981. Levels of
genetic variation in trees: influence of life history charac-
teristics. In: Conkle, M. T., ed. Symposium on isozymes
in North American forest trees and forest insects: proceed-
ings. Berkeley, CA: U.S. Department of Agricultiu-e, For-
est Service: 35-41.

Krutovskii, K. V.; PoHtov, D. V.; Altukhov, Yu. P. 1987.
Genetic variability in Siberian stone pine, Pinus sibirica
Du Tour. I. Genetic control of isozyme systems. Genetika.
23(12): 2216-2228. [In Russian]. [English translation pub-
lished by Plenvun Publishing Corp. in Soviet Genetics;
1988 June: 1568-1576].

Krutovskii, K. V.; Pohtov, D. V.; Altukhov, Yu. P. 1988.
Genetic variability in Siberian stone pine, Pinus sibirica
Du Tour. II. Level of allozyme variability in western
Sayan. Genetika. 24(1): 118-124. [In Russian]. [English
translation published by Plenum Publishing Corp. in
Soviet Genetics; 1988 July: 81-86].

Krutovskii, K V.; Pohtov, D. V.; Altukhov, Yu. P.; Milyutin,
L. I.; Kuznetsova, G. V.; Iroshnikov, A. I.; Vorobjev, V. N.;
Vorobjeva, N. A. 1989. Genetic variability in Siberian
stone pine, Pinus sibirica Du Tour. IV. Genetic diversity

and differentiation between poptilations. Genetika. 25(11):
2009-2032. [In Russian]. [EngHsh translation published
by Plenum Publishing Corp. in Soviet Genetics; 1990
May: 1343-1362].

Krutovskii, K. V.; PoHtov, D. V.; Altukhov, Yu. P. 1990. In-
terspecific genetic differentiation of Eurasian stone pines
for isoenzyme loci. Genetika. 26(4): 694-707. [In Russian].
[English translation published by Plenum Publishing
Corp. in Soviet Genetics; 1990 October: 440-450].

Krutovskii, K. V.; Politov, D. V. 1992. Study of intra- and
interspecific genetic differentiation between E\u*asian
cedar pines using isozyme loci and multidimensional
analysis. In: Symposium on molecular mechanisms of
genetic processes: proceedings. Moscow: Nauka Publish-
ers: 87-96. [In Russian].

Krutovskii, K. V.; Wagner, D. B. 1993. [Unpubhshed.] Chlo-
roplast DNA variation and phylogeny in Pinus subsec-
tions Cembrae and Strobi (subgenus Strobus). Lexington,
KY: University of Kentucky, Department of Forestry.
[Working draft].

Lanner, R. M. 1980. Avian seed dispersal as a factor in the
ecology and evolution of limber and whitebark pines. In:
6th North American Forest Biology Workshop: proceed-
ings; 1980 August 11-13; Edmonton, AB. Edmonton, AB:
University of Alberta: 15-47.

Lanner, R. M. 1982. Adaptations of whitebark pine for seed
dispersal by Clark's nutcracker. Canadian Journal of
Forest Research. 12: 391-402.

Lanner, R. M. 1990. Biology, taxonomy, evolution, and geog-
raphy of stone pines of the world. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — sjonposium
on whitebark pine ecosystems: ecology and management
of a high-mountain resource. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Research Station: 14-24.

Litvintseva, M. V. 1974. Characteristics of needle paren-
chymal cell structure in species of the group Cembrae
of the genus Pinus. Botanicheski Zhurnal. [Leningrad].
59(10): 1501-1505. [In Russian].

Little, E. L.; Critchfield, W. B. 1969. Subdivisions of the
genus Pinas (pines). Res. Rep. 1144. Washington, DC:
U.S. Department of Agriculture, Forest Service: 1-51.

Loveless, M. D.; Hamrick, J. L. 1984. Ecological determi-
nants of genetic structure in plant populations. Annual
Review of Ecology and Systematics. 15: 65-95.

Millar, C. I. 1983. A steep cline in Pinus muricata. Evolu-
tion. 32(2): 311-319.

Millar, C. I.; Kinloch, B. B. 1991. Taxonomy, phylogeny, and
coevolution of pines and their stem rusts. In: Hiratsuka, Y.;
Samoil, J. K.; Blenis, P. V.; Crane, P. E.; Laishley, B. L.,
eds. Rusts of pine. Inf. Rep. NOR-X-317. Edmonton, AB,
Canada: Northern Forestry Centre.

Nei, M. 1972. Genetic distance between populations.
American Naturahst. 106: 283-292.

Nei, M. 1975. Molecular population genetics and evolution.
New York: American Elsevier. 288 p.

Nei, M. 1977. F-statistics and analysis of gene diversity in
subdivided populations. Annals of Human Genetics. 41:
225-233.

Nei, M. 1978. Estimation of average heterozygosity and
genetic distance from a small niunber of individuals.
Genetics. 89: 583-590.

Nei, M. 1987a. Genetic distance and molecular phylogeny. In:
Ryman, N.; Utter, F., eds. Population genetics and fishery
management. Seattle, WA: University of Washington
Press: 193-223.

Nei, M. 1987b. Molecular evolutionary genetics. New York:
Columbia University Press. 512 p.

Nei, M.; Miller, J. C. 1990. A simple method for estimating
average nvunber of nucleotide substitutions within and
between populations from restriction data. Genetics. 125:
873-879.

Politov, D. V. 1989. Allozyme polymorphism, genetic differ-
entiation, and mating system in Siberian stone pine, Pinus
sibirica Du Tour. Moscow: Institute of General Genetics,
Russian Academy of Sciences. 190 p. Dissertation Thesis.
[In Russian].

Politov, D. v.; Krutovskii, K. V.; Altukhov, Yu. P. 1989.
Genetic variability in Siberian stone pine, Pinus sibirica
Du Tour. III. Linkage relationships among isozyme loci.
Genetika. 25(9): 1606-1618. [In Russian]. [Enghsh trans-
lation published by Plenum Publishing Corp. in Soviet
Genetics; 1990 March: 1053-1063].

Politov, D. v.; Krutovskii, K. V. 1990. Genetic variability
in Siberian stone pine, Pinus sibirica Du Tour. V. Analy-
sis of mating system. Genetika. 26(11): 2002-2011. [In
Russian]. [English translation published by Plenum Pub-
Hshing Corp. in Soviet Genetics; 1991 May: 1309-1316].

Politov, D. v.; Krutovskii, K V.; Altukhov, Yu. P. 1992. Iso-
zyme loci characteristics of gene banks of populations
of cedar pines. Genetika. 28(1): 93-114. [In Russian].
[English translation published by Plenum Publishing
Corp. in Soviet Genetics; 1992 July: 76-95].

Rogers, J. S. 1972. Measures of genetic similarity and ge-
netic distance. In: Studies in Genetics VII. Publ. 7213.
Austin, TX: University of Texas: 145-153.

Rohlf, F. J. 1988. NTSYS-pc: numerical taxonomy and mul-
tivariate analysis system. New York: Exeter Publishing
Ltd.

Saitou, N.; Nei, M. 1987. The neighbor-joining method: a
new method for reconstructing phylogenetic trees. Molecu-
lar Biology and Evolution. 4: 406-425.

Schuster, W. S. F.; Mitton, J. B. 1991. Relatedness within
clusters of bird-dispersed pine and the potential for kin
interactions. Heredity. 67: 41-48.

Shaw, G. R. 1914. The genus Pinus. Publ. Arnold Arboretimi
No. 5. Cambridge, IVLA: Riverside Press. 96 p.

Shurkhal, A.; Podogas, A.; Zhivotovsky, L. 1991a. Genetic
differentiation of 18 pine species on allozyme loci, genus
Pinus, subgenus Strobus and Pinus. Doklady Akademii
Nauk SSSR. 316(2): 484-488. [In Russian]. [English trans-
lation published by Plenum Publishing Corp. in Doklady
Biological Sciences; 1991].

Shurkhal, A.; Podogas, A.; Zhivotovsky, L. 1991b. Phyloge-
netic analysis of genus Pinus by allozyme loci: genetical
differentiation of subgenus. Genetika. 27(5): 1193-1205.
[In Russian]. [English translation published by Plenum
Publishing Corp. in Soviet Genetics; 1991 November].

Shurkhal, A.; Podogas, A.; Zhivotovsky, L. 1992. Allozyme
differentiation in the genus Pinus. Silvae Genetica. 41(2):
105-109.

Sneath, P. H. A.; Sokal, R. R. 1973. Numerical taxonomy.
San Francisco: W. H. Freeman: 230-234.

Steinhoff, R. J.; Joyce, D. G.; Fins, L. 1983. Isozyme varia-
tion in Pinus monticola. Canadian Journal of Forest Re-
search. 13(6): 1122-1131.

Strauss, S. H.; Doerksen, A. H. 1990. Restriction fragment
analysis of pine phylogeny. Evolution. 44(4): 1081-1096.

Strauss, S. H.; Bosquet, J.; Hipkins, V. D.; Hong, Y.-P.
1992. Biochemical and molecular genetic markers in bio-
systematic studies of forest trees. New Forests. 6: 125-158.

Swofford, D. L.; Selander, R. B. 1981. BIOSYS-1: a FOR-
TRAN program for the comprehensive analysis of elec-
trophoretic data in population genetics and systematics.
Journal of Heredity. 72: 281-283.

Szmidt, A. E. 1991. Phylogenetic and applied studies on
chloroplast genome in forest conifers. In: Hattemer, H. H.;
Fineschi, S.; Cannata, F.; Malvolti, E. V., eds. Biochemi-
cal markers in the population genetics of forest trees.
The Hague, The Netherlands: SPB Academic Publishing:
185-196.

Szmidt, A. E.; Sigurgeirsson, A.; Wang, X. R.; Hdllgren,
J. E.; Lindgren, D. 1988. Genetic relationships among
Pinus species based on chloroplast DNA pol3miorphism.
In: H^lgren, J.-E., ed. Frans Kempe Sympositun in Umeli
on molecular genetics of forest trees: proceedings; 1988;
Umed, Sweden. Ume&, Sweden: Swedish University of
Agricultural Sciences: 33-47.

Tomback, D. F. 1983. Nutcrackers and pines: coevolution or
coadaptation? In: Nitecki, N. H., ed. Coevolution. Chicago:
University of Chicago Press: 179-223.

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Coevo-
lution of whitebark pine and nutcrackers: implications for
forest regeneration. In: Schmidt, Wyman C; McDonald,
Kathy J., comps. Proceedings — symposiimi on whitebark
pine ecosystems: ecology and management of a high-
mountain resource. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
movmtain Research Station: 118-129.

Tomback, D. F.; Holtmeier, F. K.; Mattes, H.; Carsey, K. S.;
Powell, M. L. 1992. Tree clusters and growth form distri-
bution in Pinus cembra, a bird-dispersed pine. Denver,
CO: University of Colorado at Denver, Department of
Biology. [Working draft].

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird-
dispersed pines. Evolutionary Ecology. 4: 185-219.

Wheeler, N. C; Guries, R. P. 1982. Population structure,
genetic diversity and morphological variation in Pinus
contorta Dougl. Canadian Journal of Forest Research.
12(3): 595-606.

Wheeler, N. C; Guries, R. P. 1987. A quantitative measure
of introgression between lodgepole and jack pines. Cana-
dian Journal of Botany. 65: 1876-1885.

Wheeler, N. C; Guries, R. P.; O'Malley, D. M. 1983. Bio-
systematics of the genus Pinus, subsection Contortae.
Biochemical Systematics and Ecology. 11: 333-340.

Yeh, F. C. H.; Cheliak, W. M.; Dancik, B. P.; Illingworth, K;
Trust, D. C; Pryhitka, B. A. 1985. Population differentia-
tion in lodgepole pine, Pinus contorta ssp. latifolia: a dis-
criminant analysis of allozyme variation. Canadian Jour-
nal of Genetics and Cytology. 27(2): 210-218.

COEVOLUTIONAL ASPECTS OF STONE
PINES AND NUTCRACKERS

Hermann Mattes

Abstract — ^Nutcrackers (Nucifraga caryocatactes and N.
Columbiana) have developed a strong mutualistic relationship to
stone pines. Seeds of stone pines are the most important food for
at least 9 months of the year and for raising the young. In addi-
tion to special adaptations on gathering, transporting, caching,
and finding again the hoarded seeds, the whole annual cycle of
the nutcracker's life (time of breeding and moulting), its mating
system, and its habitat use are adjusted to the use of pine seeds.

Within the family of jays, magpies, and crows (Cor-
vidae) many characteristics of preadaptive value for seed
dispersal occur (Tomback and Linhart 1990; Turcek and
Kelso 1968). Three corvid species have evolved a strong
mutualistic system with pines: Clark's nutcracker (Nuci-
fraga Columbiana) and the pinyon jay (Gymnorhinus
cyanocephalus) in North America and the Exiropean nut-
cracker {Nucifraga caryocatactes) in Eurasia. Compre-
hensive studies on food regime and seed dispersal were
carried out by many authors (Bibikov 1948; Crocq 1990;
Hayashida 1982, these proceedings; Marzlviif and Balda
1992; Mattes 1978; Reijmers 1959; Saito 1983; Swanberg
1951; Tomback 1977; Vander Wall and Balda 1977, 1981).
Coevolution in corvids and pines does not only refer to
feeding and caching behavior, but the whole life cycle is
adjusted to the use of the special food resource. This pa-
per describes the process for the European nutcracker.

THE EUROPEAN NUTCRACKER

The annual life cycle (fig. 1) of the nutcracker is best
looked at fi'om August onward. Unripe seeds of the Swiss
stone pine (cembra pine, Pinus cemhra) are eaten by the
nutcracker throughout August or even in July; thus, the
bird will not miss the point when the seeds are fully de-
veloped. At this time the resin flow in the cone stops, and
cone scales get loose. Now the seeds can be easily taken
out of the cone, and the nutcracker begins gathering and
caching seeds for winter supply. Pine seed gathering will
last until all seeds have been harvested and almost no
cones are left on the trees. Only twice within 19 years of
investigation in the Engadine a cone crop lasted until the
beginning of the following spring.

In the Alps, nutcrackers carry seeds of cembra pine over
distances up to 15 km (including differences in altitude of
at least 700 m) back to their territories. They carry 45

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Hermann Mattes is Professor of Landscape Ecology and Geography,
Department of Geography, University of Miinster, Germany.

pine seeds per flight on an average; however, a maximum
load of 134 cembra pine seeds has been found in the sub-
lingual pouch of a nutcracker. Main morphological and
ethological adaptations for seed use are:

• a long and pointed bill (fig. 2) that enables the bird

to take seeds out of the cones as one would with twee-
zers (most other corvids have a hooked bill);

• a rist in the lower mandible allows the bird to fix and
crack pine seeds, supported by a short-cut tongue;

• the subhngual pouch allows carrying a large quantity
of seeds (fig. 3) without disturbing other functions of
the bill;

• ability to test seed quality by means of seed color and
resonance (by so-called bill-clicking).

SEED CACHES

The cached seeds provide the main food source for the
nutcracker from November to April, and to a lesser degree
until July. A single nutcracker stores more than 25,000
seeds. As yet it has not been clearly shown how the nut-
cracker can retrieve its seed caches. However, as can be
concluded fi'om observations in the field and in the labora-
tory, nutcrackers apparently remember each single cache
(Tomback 1982, and others). Further, there are no spe-
cific features of caches, although nutcrackers obviously
prefer to deposit seeds near border lines and any changes
of surface structure. Experiments with the Clark's nut-
cracker by Kamil and Balda (1988) and Vander Wall
(1982) showed that nutcrackers orient (navigate) opti-
cally, as expected, and additionally with a magnetic sense.
Understanding cache findings, however, requires more de-
tailed information, as nutcrackers detect caches even un-
der a thick cover of snow. Balda and Kamil (1989) sug-
gest an overall fuzzy optical navigation for finding caches.
This wovild allow the birds to locate a cache even if there
were distinctive changes, for example, in height of snow
cover, broken branches, or fallen trees. It is still an open
question how the bird can find caches under a snow cover
of 20 cm and higher, getting very directly to the hoarded
seeds. Snow tunnels of 130 cm (Burckhardt 1958) and
about 300 cm (R. Stem, personal communication) were
reported, and — what is most surprising — the nutcrackers
did find the cached seeds. Crocq (1990) found the nut-
crackers leaving their subalpine habitat when snow cover
reached 170 cm on imdistixrbed surfaces.

It shoiild be emphasized that nutcrackers dig success-
fully for their caches throughout winter imtil early spring.
That guarantees sufficient food for the brood during that
difficult time of the year. Nutcrackers must remember
not only the sites of the caches but additionally those that
already have been used. This doubles the efforts of

twmm

Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

Month

Figure 1 — Annual activities of the nutcracker in the Alps for years with a medium cone crop. 1 -3, feeding:
1 , seed harvest and transport; 2, use of cached seeds; 3, consumption of animal food, for example, spi-
ders, insects; 4, winter period; 5-7 breeding period: 5, courtship and nest building; 6, time with clutch and
nestlings; 7, family groups with fledged young; 8-9, moulting: 8, moult of adults; 9, moult of fledglings;
10-11, movements: 10, dismigration; 11, migration.

memory. There is some evidence that nutcrackers use
caches from the penultimate seed crop, which may be im-
portant in case a cone crop fails. In years of small seed
crops, nutcrackers often recache the seeds. Temporary
caches are made in cases of strong competition for seeds
(when great nimabers of nutcrackers and squirrels concen-
trate on a few cone-bearing trees). In temporary caches,
nutcrackers mostly hide cones.

BREEDING AND HABITAT

Nutcrackers start breeding in March (fig. 4). At this
time mean daily temperature is under 0 °C. Early breed-
ing is necessary because the next generation has to be
ready for harvesting seeds in August to ensure the winter
food supply. Breeding takes about 18 days, the nestlings
fledge when 24 to 28 days old, and there is an extraordi-
narily long time of parental care for the fledglings — until
they are about 100 days old (Oberauer 1991; Rudat 1978).
Extensive care by the parents is needed because of scarcity

Figure 2 — Inside the nutcracker's bill: short
split tongue, underneath opening of sublin-
gual pouch with peanut in the entrance part,
rist in the lower mandible.

Figure 3 — Nutcracker with filled sublingual
pouch. Morteratsch, Engadin; September
1975.

of food up to mid-June, and to provide training for caching
and digging for seeds.

Courtship behavior starts in the end of February or
early March. Synchronizing with the partner is of ex-
treme importance for successful breeding. By February
the nutcrackers' so-called ceremonial gatherings begin
and then can be frequently observed (Swanberg 1956).
Up to 12 or more nutcrackers from neighboring and some-
times far territories meet in conspicuous noisy groups. An
individually marked female nutcracker that had spent the
winter outside its territory came back late — March 17. I
discovered her fighting for the territory while a ceremo-
nial gathering was going on. She succeeded in reoccupy-
ing her former place, and raised two nestlings after hav-
ing a three-egg clutch with one unfertile and half-sized
egg. Neighbors know each other individually. So, except
in artifically created feeding places, fights are very rare.

Nutcracker territories are organized in a different way
than those of other passerines. Only the area very close
to the nest is defended against intruders and thus can be
called "territory" in the usual sense. In a broader sense,
the territory comprises the area of food caches and is
about 5 to 12 ha large according to habitat. These "terri-
tories" can overlap and are not defended because cached
seeds are only accessible to the very same nutcracker. In
addition, all places regularly visited for seed gathering
and transport should be regarded as home range. Habi-
tats of the nutcracker are characterized by evergreen co-
nifers. In such forests sheltered nesting places are avail-
able, and because snow cover is low under dense tree
canopies cached seeds can be easily recovered.

Nutcrackers are adapted in many ways for breeding
under winter conditions. The nest is heavily built with a
layer of dry decayed wood for insulation against low tem-
perature. Beginning with the first egg, the parents must
warm the clutch without long interruptions. Breeding
starts after the last egg has been laid, and egg tempera-
ture is raised from about 25 °C to breeding conditions of
about 38 °C. Different from most corvids, both adults
breed. The male develops a large breeding patch, too.
The female's breeding time is somewhat higher during the
day, and it also stays on the nest at night. The nestlings
are fed with pine seeds from their first day of life. Addi-
tional food consists of spiders, and, if available, insects.
Nutrient analysis in seeds of whitebark pine {Pinus
albicaulis) by Lanner and Gilbert (these proceedings)
showed a high nutritive quality in caloric value, mineral
nutrients, fatty acids, and amino acids. Should this apply
also to cembra pine seeds, it is easy to understand why
first-day nestlings can live mainly on a seed diet.

For successful overwintering and reproduction, suffi-
cient numbers of seed caches must be available to both
parents. First-year birds dismigrate in August and estab-
lish their own territory when seed caching starts. First-
year birds that are not fully developed until August will
have a minor chance to survive winter. A few first-year
birds leave their native area in September and migrate in
southern directions (Mattes and Jenni 1984). Adult nut-
crackers are highly sedentary. Nonbreeding individuals
might have action ranges within or between the territo-
ries of breeding birds. This popiilation reserve can com-
pensate for winter losses of adult birds.

In the "Stazer Wald" I observed a third bird helping a
pair build their nest; unfortunately, no further observa-
tion was possible. Similar to several jay species, helping
in breeding may be an evolutional feature in the nut-
cracker (Woolfenden and Fitzpatrick 1984). Nutcrackers
live monogamously for life, as can be concluded from indi-
vidually marked birds (Mattes 1978; Swanberg 1956). In

Figure 4 — A nutcracker breeding, the nest
completely covered by snow. March 1 9,
1975, Engadin.

the Engadine, the mortality rate in adult birds was 84 per-
cent (Mattes, in preparation) on an average. The oldest
birds found were at least 15. Keeping territories and part-
ners for life supports synchronization of the parents, and
individual knowledge of neighbors minimizes social com-
petition. Social hierarchy is evident and most conspicu-
ous at artifical feeding places. Females take over the so-
cial rank of their mates (Grabher-Meyer 1991).

In passerines, moult normally takes place after the
breeding season. Prolonged care of the fledglings and
beginning of seed hoarding in August would not leave
time for movdting in late summer. Actually, moult starts
in March with change of remiges and tail feathers, and
ends in July when body feathers are changed. Moult com-
pletely parallels breeding season; the double physiological
effort causes a very long moulting period. Body feathers
are moulted relatively quickly during the warmest part of
the year.

REACTION TO CHANGES

Predation on pine seeds has influenced the whole life
cycle of the nutcracker. Many features common to the
corvids are particularly developed in the nutcracker. Still
there is a broad and flexible reaction to environmental
changes. This is most apparent during the invasions of
Siberian nutcrackers (N. c. macrorhynchos) into western
Europe. Diet and feeding as well as habitat use vary
widely. Nevertheless, most of the invading birds very
likely die. Some, however, return to their breeding ground
either in the current autumn or in the next spring. Nimaer-
ous birds ringed in the Aland archipelago (Finland) in
August 1968 were recaptured a few weeks later in the
East; three birds were found beyond the Urals at a maxi-
mum distance of almost 3,000 km (Zink 1981). Invaders
stayed at wintering places in Bielefeld, Germany, when
Siberian pine seeds (Pinus sibirica) were offered ad
libidimi (Conrads and Balda 1979).

Outside the natural range of cembra pine (Central Alps,
Tatra, and a few locations in the Carpathian mountains)
the European nutcracker normally feeds on hazel nuts
iCorylus avellana). Slight, but statistically significant,
differences in bill size are apparent in central-alpine
populations. Nutcrackers from the Engadine and migrat-
ing birds at Col de Bretolet in southwestern Switzerland
show a narrower lower mandible and a longer bill com-
pared with populations from southern Germany and from
northern Europe (Mattes 1978; Mattes and Jenni 1984).
Also, there are distinct differences in bill size in popula-
tions of A^. c. macella from Szetchuan and from Sikkim
(Mattes, unpublished). Generally, there seems to be a
correlation of width of lower mandible and size of main
food. The distance between the branches of the lower
mandible limits the size of the entrance to the sublingual
pouch. As the main food of most Asian populations of the
nutcracker is not yet sufficiently known, bill size and bill
structure cannot yet be fully explained. Furthermore, bill
size in some subspecies may reflect feeding conditions at
former times.

REFERENCES

Balda, Russel P.; Kamil, Alan C. 1989. A comparative

study of cache recovery by three corvid species. Animal

Behavior. 38: 486-495.
Bibikov, D. I. 1948. On the ecology of the nutcracker.

Trudy Pechoro-Ilychkoge Gosud. Zapovednik. 4: 89-112.
Biu-ckhardt, D. 1958. Vom Vorratanlegen des Tannen-

hahers. Biindnerwald. 11:102-114.
Conrads, Klaus; Balda, Russell P. 1979. Uberwinter-

imgschancen Sibirischer Tannenhaher {Nucifraga

caryocatactes macrorhynchos) im Invasionsgebiet.

Bericht des Naturwissenschaftlichen Vereins Bielefeld.

24: 115-137.

Crocq, Claude. 1990. Le casse-noix mouchete {Nucifraga
caryocatactes). Paris: Lechevalier/R. Chabaud. 326 p.

Grabher-Meyer, Arno. 1991. Beobachtungen zur Brut-
biologie einer kleinen Population des Tannenhahers
{Nucifraga caryocatactes) bei Praxmar, Sellrain.
Diplomarbeit, Universitat Innsbruck, Austria.

Kamil, Alan C; Balda, Russell P. 1988. Spatial memory
in Clark's nutcracker {Nucifraga columbiana). Acta XIX
Congressus Internat. Ornithol., Ottawa 1986. Vol. II:
2100-2111.

Lanner, Ronald M. 1982. Adaptations of whitebark pine
for seed dispersal by Clark's nutcracker. Canadian
Journal of Forest Research. 12(2): 391-402.

Marzluff, John M.; Balda, Russ P. 1992. The pinyon jay:
behavioral ecology of a colonial and cooperative corvid.
London: Poyser. 317 p.

Mattes, Hermann R. 1978. Der Tannenhaher im Engadin.
Studien zu seiner Okologie und Funktion im Arven-
wald. Miinstersche Geographische Arbeiten 2.
Paderbom: Schoning. 87 p.

Mattes, Hermann R.; Jenni, Lukas. 1984. Ortstreue und
Zugbewegungen des Tannenhahers {Nucifraga caryo-
catactes) im Alpenraima und am Randecker Maar/
Schwabische Alb. Ornithologischer Beobachter. 81:
303-315.

Oberauer, Eva. 1991. Zur Postembryonalentwicklimg des
Tannenhahers {Nucifraga caryocatactes L.). Beobach-
tungen im Alpenzoo Innsbruck. Diplomarbeit, Uni-
versitat Innsbruck, Austria. 140 p.

Reijmers, N. F. 1959. Birds of the cedar-pine forests of
south-central Siberia and their role in the life of the
cedar-pine. Trudy biolog. Inst, sibirsk. Otdel. Akad.
Nauk SSSR. 5: 121-166.

Rudat, Volker; Rudat, Wolf. 1978. Zum Verhalten von
Tannenhaherfamilien {Nucifraga caryocatactes L.) in
der Zeit vom Ausfliegen der Jungen bis zu deren
Selbstandigwerden. Zoologisches Jahrbuch fur
Systematik. 105: 386-398.

Saito, Shin-ichiro. 1983. On the relations of the caching
by animals on the seed germination of Japanese stone
pine, Pinus pumila Regel. Bulletin of the Shiretoko
Museum. 5: 23-40.

Swanberg, Peer O. 1951. Food storage, territory and song
in the thickbilled nutcracker. Acta X Congressus
Internat. Ornitholog. Uppsala 1950: 545-554.

Swanberg, Peer O. 1956. Territory in the thickbilled nut-
cracker {Nucifraga caryocatactes). Ibis. 98: 412-419.

Tomback, Diana F. 1977. Foraging strategies of Clark's

nutcracker. The Living Bird. 17: 123-161.
Tomback, Diana F. 1982. How nutcrackers find their seed

stores. The Condor. 82: 10-19.
Tomback, Diana F.; Linhart, Yan B. 1988. The evolution

of bird-dispersed pines. Evolutionary Ecology. 4:

185-219.

Vander Wall, Stephen B. 1982. An experimental analysis
of cache recovery in Clark's nutcracker. Animal
Behavior. 30: 84-94.

Vander Wall, Stephen B.; Balda, Russ P. 1977. Coadapta-
tions of the Clark's nutcracker and the pinyon pine for

efficient seed har\'est and dispersal. Ecological Mono-
graphs. 47(1): 27-37.
Vander Wall, Stephen B.; Balda, Russ P. 1981. Ecology
and evolution of food storage behavior in conifer-
seed-caching Corvids. Zeitschrift fiir Tierpsychologie.
56: 217-242.

Woolfenden, D. E.; Fitzpatrick, W.-J. 1984, The Florida
scrub jay. Princeton, NJ: Princeton University Press.

Zink, Gerhard. 1981. Der Zug europaischer Zugvogel. 3.
Liefenmg. Moggingen: Vogelzug-Verlag.

ALLOZYME POLYMORPHISM,
HETEROZYGOSITY, AND MATING
SYSTEM OF STONE PINES

Dmitri V. Politov
Konstantin V. Krutovskii

Abstract— Population genetic structure and the mating system of
stone pines (genus Pinus, subsection Cembrae) have been studied
using isozymes. Observed were a tendency to heterozygote defi-
ciency among embryos (obviously, a consequence of self-pollination
with the estimated fi-equency of 0.11-0.30) and a slight heterozy-
gote excess in adult trees (probably caused by selection against
inbred progeny and balanced selection).

Electrophoretic analysis of allozjone polymorphism is
one of the most useful methods for study of genetic and evo-
lutionary processes in natural populations and taxonomic
and phylogenetic relationships of forest tree species (Adams
and others 1992; Fineschi and others 1991; MuUer-Starck
and Ziehe 1991). Conifers are particularly attractive, from
a genetic point of view, because of their haploid megaga-
metophyte (endosperm). Electrophoresis of both haploid
and diploid (embryo) seed tissues permits us to distinguish
maternal and paternal contributions into the progeny geno-
type. This makes it possible to estimate allele frequencies
separately in both male and female gametes pools and to
study segregating ratios directly by analyzing endosperm
haplotypes, which correspond to haplotypes of female ga-
metes. The possibility of observing meiotic products di-
rectly through study of haploid endosperms simplifies analy-
sis of allozyme inheritance, genetic Unkage of isozyme loci,
and the mating system of conifers.

In contrast to many widespread European and North
American conifer species, stone pines have been studied
using isozymes only recently (Groncharenko and others 1988;
Krutovskii and Politov 1992; Krutovskii and others 1987,
1988, 1989, 1990; Podogas and others 1991; Politov 1989;
PoUtov and Krutovskii 1990; Politov and others 1989, 1992;
Szmidt 1982). Meanwhile, the need for such data is appar-
ent both for practical forestry applications and for purposes
of genetic diversity conservation. As a major forest-forming
tree species, stone pines determine, to a large extent, the
state of ecosystems over extensive areas. Their genetic vul-
nerability thus threatens the existence of a unique forest
type.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Dmitri V. Politov is Research Scientist and Konstantin V. Krutovskii
is Senior Research Scientist, Laboratory of Population Genetics, Russian
Academy of Sciences, Gubkin Str. 3, 117809 GSP-1 Moscow, Russia.

Four of the five stone pine species traditionally united
in subsection Cembrae of genus Pinus (Critchfield and
Little 1966) are found in Russia and neighboring countries:
Siberian stone pine (Pinus sibirica Du Tour), European or
Swiss stone pine (P. cembra L.), Korean pine {P. koraiensis
Siebold et Zucc), and mountain or Japanese stone pine
(P. pumila [Pall.] Kegel). All of them are very important
economically.

This paper reports the main results of long-term popula-
tion genetic studies on these species using isozyme analy-
sis. Some population structure parameters also have been
estimated for one population of whitebark pine (P. albicaulis
Engelm.), the only representative of stone pines in the
Western Hemisphere.

MATERIALS AND METHODS

Sampling Sites— Seeds collected in natural populations
served as a source of laboratory material. Pinus sibirica
seeds were collected from 11 localities. Samples desig-
nated as Filin Klyuch (F), Mutnaya Rechka (M), Maljd
Kebezh (MK), Sobach'ya Rechka (SR), and Listvyanka (L)
originate from the Western Sayan (Ermakovskii District,
Krasnoyarsk Territory); Yayl'u (YA) from the Altfd Moun-
tains (Gomo-Altaiskaya Autonomous Region); Smokotino
(S), Zorkal'tsevo-1 (Z-1), and Zorkal'tsevo-3 (Z-3) from Tom-
skaya Region; Turukhan (TU) from the north of Krasnoy-
arsk Territory (Turukhanskii District); and N'alino (NYA)
from the Khanty-Mansiiskii Autonomous District (Tjoimen-
skaya Region). Besides, one locality of P. cembra, Ust'-
Chornaya (UCH) from the Eastern Carpathians (Ukraine);
three localities of P. pumila from the north of Kamchatka
Peninsula designated as Grustnyi (GR), Uttyveem (UT),
and Kamenskoye (K); and three localities of P. koraiensis
from Khabarovsk Territory designated Malokhekhtsirskoye
(MKH) and Khor (KHO), and from Primorskii Territory
(Sikhote-Alin' (SA)) have been also studied. The sample
of P. albicaulis originates from the Livingstone Falls (LF)
population (southwestern Alberta, Canada).

Electrophoresis and Data Processing— No less than
seven endosperms and embryos per tree have been studied
using starch gel electrophoresis. Totally, 536 trees of five
species (from 11 to 75 trees per population) and more than
3,500 embryos (7 to 20 per tree), as well as two bulked sam-
ples of seeds representing a large number of trees (TU and
NYA), have been analyzed. Specimen preparation, buffer
systems used, genetic interpretation of zymograms, and des-
ignations of £dlozymes, alleles, and loci have been described
elsewhere (Krutovskii and others 1987; Politov 1989).

Genotypes of particular trees were inferred from segre-
gation of alleles among their endosperms. Then, allele and
genot5T)e frequencies in samples of both mature trees and
embryos, as well as in the effective pollen pool (paternal
alleles of embryos), were estimated. Embryo tissue elec-
trophoresis does not give high resolution for all zones of
activity and, therefore, calculations based on embryo sam-
ples have included a smaller nimiber of loci. Besides, some
activity zones in individual species could not be reliably in-
terpreted or were invariant. Because of this, specific sets
of loci used for particular purposes in these species may dif-
fer somewhat. More detailed description of statistic proce-
dures can be fovmd elsewhere (Krutovskii and others 1989;
Politov 1989; PoHtov and Krutovskii 1990; Politov and oth-
ers 1989).

GENETIC CONTROL OF ISOZYMES
AND LINKAGE OF LOCI

We have described inheritance of 11 isozyme systems
of stone pines in our previous publications (Krutovskii and
others 1987; Politov 1989) and foimd that they are controlled
by at least 26 loci. However, only 21 most reliable loci with
unambiguous genetic control have been included in this
study (table 1). Segregation ratios of allelic variants of in-
dividual allozjnne loci observed among endosperms of het-
erozygous trees correspond, with several exceptions, to the
expected ones (1:1).

Analysis of joint allelic segregation for pairwise loci com-
binations in Pinus sibirica has revealed genetic linkage
of loci Adh-1 and Lap-3 (with the mean recombination fi^e-
quency 0.275), Adh-1 and Pgi-2 (0.301), and Fe-2 and Lap-2
(0.309). These loci belong to two linkage groups: Lap-3 —
Adh-1— Pgi-2 and Lap-2— Fe-2 (Politov and others 1989).
Comparison of these data with those available in literature
shows that stone pines, as well as other representatives of
the family Pi/iacme, are characterized by extreme conser-
vatism of their genome, since the same loci combinations
with nearly the same values of recombination fi-equency
have been described not only in the closely related P. albi-
caulis (Furnier and others 1986), but also in many hard
pines (see, for example, Szmidt and Muona 1989). Data
presented here are apparently the first attempt of genetic
mapping in one of the stone pines. However, one can as-
sume that the linkage groups revealed in P. sibirica will
also be fotmd in other related species. Information of this
kind could be useful for breeding programs, for analysis
of association of genetic markers with economically impor-
tant characters, and for mating system parameters estima-
tion by multilocus methods.

POPULATION GENETIC STRUCTURE

Levels of Intrapopulation Variability — ^According
to the gene diversity parameters estimated in stone pine
populations using isozjone loci, the level of genetic vari-
ability appears to be rather high (table 2). There are al-
most no differences between observed heterozygosities and
those expected fi'om corresponding Hardy- Weinberg pro-
portions. The values of the latter for individual species
are: for P. sibirica— from 0.140 to 0.176, P. cembra— 0.109,

Table 1 — Enzymes and loci analyzed

Total
number

Enzyme (abbreviation)

of loci

Scored loci

Alcohol dehydrogenase (ADH)

Adh-1 , Adh-2

DIaphorase (DIA)

Dia-1 , Dia-2

Fluorescent esterase (FE)

Fe-2

Glutamate dehydrogenase (GDH)

Gdh

G 1 utam atoxaloacetate

Got-1 , Got-2, Got-3

transaminase (GOT)

Isocitrate dehydrogenase (IDH)

Idh

Leucine aminopeptidase (LAP)

LaD-2 LaD-3

Malate dehydrogenase (MDH)

Mdh-1, Mdh-2, Mdh-3

Mdh-4

Phosphoglucose isomerase (PGI)

Pgi-2

Phosphoglucomutase (PGM)

Pgm-1, Pgm-2

Shikimate dehydrogenase (SKDH)

Skdh-1 , Skdh-2

P. pumila — from 0.239 to 0.258, P. koraiensis — from 0.113
to 0.148, P. albicaulis— 0.204.

Such a high amoimt of intrapopulation variability is char-
acteristic of conifers. We have summarized data on allozyme
diversity of 30 pine species using available references
(Politov and others 1992). The mean number of alleles per
locus (A) and the expected heterozygosity (Hg), averaged for
all pine species, equalled 2.08 and 0.169, correspondingly.
Comparable values have been found for spruces (A = 2.2,
Hg = 0.183), larches (A = 1.6, H^ = 0.091), and representa-
tives of other less-studied genera.

Factors supposedly promoting a high level of genetic diver-
sity in stone pine populations include such ones, common
for forest trees, as predominantly outcrossing mating sys-
tem, wide distribution, large effective size of population,
and high longevity (Hamrick and Godt 1989; Ledig 1986).
More specific evolutionary factors, such as drastic reduc-
tion of population size ("bottleneck effect"), may also affect
the variability level. Although the ranges of stone pines,
both in Eurasia and in North America, were reduced diu--
ing the Quaternary glacials (and also during the global
warming), preservation of extensive refugiimis permitted
avoidance of substantial genetic depauperization. This
concerns Pinus cembra to a much less degree than other
stone pines.

According to some evidence, Swiss stone pine imderwent
a sharp reduction in its range and population size during the
Holocene caused by the global warming (Bobrov 1978 and
references therein). A small effective number is typical of
most populations of this species at present as well. Swiss
stone pine exists nowadays in the Alps and the Carpathians
as small isolated stands in which processes of genetic drift
do not allow aUozyine variability to be maintained on a high
level. This is a most probable cause for a lower level of het-
erozygosity (0.109) and the percentage of polymorphic loci
(26.3 percent) of P. cembra, as compared to other species.
However, it should be noted, that the only sample of this spe-
cies studied by us represents a small marginal stand from
the eastern Carpathians, while central populations may
appear to be more variable. The limited nimaber of loci
studied in the only work (Szmidt 1982) dealing with allo-
zyme polymorphism in Swiss stone pine over the species

Table 2 — Levels of allozyme variability in stone pine populations

Species and Number Number

population names

of loci

of trees

Ho'

He'

Dim 1^ ^ihifS^o
ninus SiOiriCa

Filin Klyuch^ (F«)

53.0

1.8

47.4

0.155

0.152

Mutnaya Rechka (M)

75.0

1.9

47.4

.137

.140

Maiyi Kebezh (MK)

15.0

1.8

52.6

.158

.153

Sobach'ya Rechka (SR)

12.0

1.7

42.1

.145

.173

Listvyanka (L)

41.0

1.9

42.1

.163

.151

Yayl'u (YA)

43.0

1.9

47.4

.153

.154

Smokotino (S)

43.0

1.7

42.1

.168

.161

Zorkal'tsevo-I (Z-1)

34.0

1.7

47.4

.198

.176

Zorkartsevo-3 (Z-3)

34.0

1.7

47.4

.184

.165

Mean

38.9

1.8

46.2

.162

.158

UsfChornaya (UCH)

15.0

1.5

26.3

.128

.109

Pinus pumila

Grustnyi (GR)

54.0

2.3

50.0

.230

.239

Uttyveem (UT)

55.0

2.5

60.0

.243

.251

Kamenskoye (K)

55.0

2.3

70.0

.269

.258

Mean

54.7

2.4

60.0

.247

.249

Pinus koraiensis

Malokhekhtsirskoye (MKH)

30.0

1.6

47.1

.147

.132

Khor (KHO)

16.0

1.6

47.1

.162

.148

OlKllOie-Mlin ^OA;

1 7
1 /

1 1 .u

1 .O

.Mo

Mean

19.0

1.6

43.2

.137

.131

Pinus albicaulis

Livingstone Falls (LF)

52.0

1.9

56.3

.213

.204

'Mean number of alleles per locus.

^Mean percent of loci polymorphic (locus has been considered polymorphic if the frequency of the most com-
mon allele has not exceeded 95 percent).
^Mean heterozygosity observed.
"Mean heterozygosity expected.
^See text for origins of populations.
^Abbreviated denomination of population.

range does not allow any conclusions to be made about the
variability level.

Heterozygosity of Korean pine (mean 0.131) is slightly
higher. As to Pinus pumila, this species shows the highest
values of H among stone pines (0.249), close to the maxi-
mum ones known for conifers (Politov and others 1992).
One cannot exclude the fact that such a high variability
is very important for adaptation of this species to a wide
spectnmti of environmental conditions, including those of
permafrost regions, and drastic seasonal climatic changes.
Nearly the same heterozygosity level (0.204) was found in
the only studied population of whitebark pine from the
north moimtain area.

Genotype Distribution in Adult Trees and
Embryos — Comparison of observed genotype distributions
with those expected from the corresponding Hardy- Weinberg
proportions is commonly used in population genetics to re-
veal and identify various factors affecting genetic structure
of populations. In stone pines such analysis has not re-
vealed pronounced deviations from Hardy- Weinberg equi-
librium for mature trees in any samples.

Statistically significant excess or deficiency of heterozy-
gotes have not been observed for allozyme loci studied in
stone pine populations. Nevertheless, the sign criterion
showed that there were significantly more cases of a slight
heterozygote excess (Politov and others 1992), which may
mean existence of a tendency toward systematic excess of
heterozygotes of alloz3mie loci among adult trees. Figure 1
shows mean values of Wright's fixation index (or "inbreed-
ing coefficient," F = l-H/Hg, where H^ and Hg are mean
observed and expected heterozygosities for all loci studied
in the species). Predominance of slightly negative or almost
zero F values indicates absence of significant heterozygote
deficiency in the reproductive part of the popvilations. Spe-
cies mean values were: for P. sibirica, -0.025; for P. cembra,
-0.174; for P. koraiensis, -0.046. F values for P. pumila
were slightly positive in two samples and negative in the
third one (mean F = 0.008).

In most embryo samples, on the contrary, a slight defi-
ciency of heterozygotes was regularly observed (Politov and
others 1992). Figure 2 presents mean F values in embryos
and adult trees calculated for the same loci. In embryos

0.1
0.05

F 0

-0.05
-0.1
-0.15

A.
'1

niri

Populations: F M MK SR L YA S Z-1 Z-3 MEAN

GR UT K MEAN MKH SA MEAN

J KHO
I

P. koraiensis

P. sibirica P. cembra P. pumila

Figure 1 — Observed fixation index ("inbreeding co-
efficient") values (F) in samples of adult trees. All
studied loci used. (For full names and origins of
populations see text and table 2).

the values are positive in nearly all samples (except one),
whereas mature trees display mainly negative F values.

Probable Causes of Heterozygote Deficiency —

most probable cause for heterozygote deficiency at early de-
velopmental stages of stone pines may be inbreeding. The
species studied, as well as other conifers, have a mixed mat-
ing system (Brown and others 1985; Politov and Krutovskii
1990 and references therein) characterized by a random
outcrossing with a certain percentage of selfing. Heterozy-
gote deficiency may theoretically be caused by both self-
pollination and other consanguineous matings (Ritland and
El-Kassaby 1985; Shaw and others 1981). One cannot also
exclude the action of natural selection against heterozy-
gotes at a certain developmental stage.

Much evidence has been reported demonstrating hetero-
zygote deficiency among embryos in populations of conifers
(summarized by Bush and Smouse, 1992). In some rare
cases deficiency is revealed in mature trees. The authors
explained this phenomenon by inbreeding and, particularly,
by selfing.

Mating System Analysis — ^The fact that selfing is most
commonly regarded as a factor causing inbreeding in conifer
species has good reasons. Coincidence, in general, of polli-
nation with pistil receptivity of the same tree and absence of
effective self-sterility mechanisms make this process highly
probable. In order to test the h3T)othesis that partial selfing
is the main cause of heterozygote deficiency in stone pines,
we have applied methods of mating system parameters es-
timation based on allozyme data and mixed mating system
theory (Brown and others 1985). Use of isozymes and pow-
erful mathematical procedures (Expectation-maximization
method, Newton-Raphson method, bootstrap, etc.) have
made it possible to obtain quantitative estimates of selfing
and outcrossing rates in natural conifer populations. For
different species percentage of selfing varies between 2 to 3

and 20 to 30 percent (see Brown and others 1985; Politov
and Krutovskii 1990, for review). For stone pines such
estimates have not been made until recently. Here we
present data obtained on three stone pine species.

The results of mating system analysis of Siberian stone
pine have been published earlier (Politov and Krutovskii
1990). If we consider a randomly outcrossing population
at eqmlibrium, absence of pre- and postzygotical selections,
sexual symmetry, separated generations, etc., the expected
inbreeding coefficient Fg can be derived from the equation
Fg = (l-^)/(l+^), where t is the outcrossing rate estimated
for this species. In our study the expected inbreeding coef-
ficient Fg has turned out to be very close to the actually
observed one (F^) [F„ = iK~R^)fH^, where Hg and H^ are,
correspondingly, the expected from Hardy- Weinberg pro-
portions and the directly observed heterozygosities] in em-
bryo samples. Thus, selfing (1-t) whose estimated rate is,
on the average, as high as 15 percent (from 2 to 27 in indi-
vidual populations) is most likely to be one of the main fac-
tors of heterozygote deficiency occurrence among embryos
of this species.

Available data for three stone pine species are presented
in table 3. Estimation of outcrossing rates was made based
on four to eight polymorphic loci and using two different
computation procedures — variants of both single- and mtdti-
locus methods (Ritland and El-Kassaby 1985; Ritland and
Jain 1981). Both procedures have been realized by Ritland
(1990) in his original computer program MLT. The multi-
locus estimation (which has more detective power than the
single-locus one) gives for Pinus sibirica populations values
ranging from 0.846 to 0.980 (mean 0.894), and for P. korai-
ensis values fi:'om 0.920 to 1.034 (mean 0.974). The out-
crossing rate foimd in P. cembra was substantially lower

P. sibirica

P. cembra i_

Q Adult trees
I Embryos

Figure 2 — Observed fixation index, "inbreeding
coefficient" (F^), values in samples of adult trees
(striped bars) and embryos (filled bars). Only
eight common loci used. Fg is expected inbreed-
ing coefficient [Fg = {^-tJ^{^+tJ]. S is selfing
rate (s = l-f^, see table 3). (For full names and
origins of populations see text and table 2).

P. lioraiensis

Table 3 — Multilocus (f J and single-locus {Q estimates of outcrossing
rate in stone pine populations

Species and Number Outcrossing rate

population names of loci t„ (s.e.) (s.e.)

Pinus sibirica

Filin Klyuch-85^

8.0

0.817

(0.031)

0.788

(0.030)

Filin Klyuch-86

8.0

.892

(.042)

.884

(.052)

Mutnaya Rechka-86

8.0

.980

(.019)

.959

(.048)

iviuiiidyct nuoiirVd 0/

8.0

.863

\.\JOO)

.0^0

1 n'^fi^

Maiyi Kebezh

8.0

.960

(.072)

.933

(.060)

Sobach'ya Rechka

8.0

.846

(.125)

.834

(.105)

^lOlVyCII IrVCl

8.0

.912

(.031)

fl71

Yayl'u

8.0

.855

(.045)

.776

(.043)

Smokotino

7.0

.929

(.033)

.886

(.037)

Mean

7.9

.894

(.057)

.862

(.054)

Pinus cembra

Ust'-Chornaya

4.0

.686

(.025)

.707

(.045)

Pinus

koraiensis

Malokhekhtsirskoye

6.0

.967

(.062)

.914

(.052)

Khor

7.0

.920

(.037)

.929

(.066)

Sikhote-Alin'

6.0

1.034

(.070)

.964

(.031)

Mean

6.3

.974

(.058)

.936

(.051)

'Number means year of sampling of material; see text for origins of
populations.

(0.686), but we regard this value only as a preliminary one
because of the limited sample size and number of loci stud-
ied. Nevertheless, it may also be the real level of selfing,
probably due to the limited nimiber of trees in the studied
locality and a low stand density.

Shaw and others (1981) note that we can conclude the role
of inbreeding components other than selfing per se in con-
sanguineous matings (for example, mating between closely
related neighbor trees in a population with family structure
or assortative/selective pollination/fertilization between
genetically similar trees) from the difference between the
values of multi- and single-locus estimates. In oiu* study of
stone pines these differences, for nearly all samples, appesir
to be negligible (the difference, in general, does not exceed
0.04). This implies a rather weak effect of mating between
relatives or, in other words, absence of microsubdivision of
the population.

Fumier and others (1987) found that genetic and physical
distances between clumps of Pinus albicaulis fail to corre-
late. The authors related this fact to the effective "mixing"
of seeds in the process of their dispersal by the Clark's nut-
cracker {Nucifraga columbiana Engelm). We can expect to
observe randomization of such kind (more closely spaced
seedlings or clumps are not more related ones) in any other
pine species depending on the European nutcracker (AT. cary-
ocatactes L.) in their dispersal, since the effectiveness of nut-
crackers as seed dispersers has been repeatedly proved by
numerous investigations (Lanner 1980, 1990; Tomback
and Linhart 1991).

Selection Against Inbred Progeny — What may cause
"disappearance" of heterozygote deficiency observed among
embryos, as compared to adult trees of the population

(figs. 1 and 2)? This can, most likely, be the result of selec-
tion against inbred progeny formed by self-pollination, as
well as selection in favor of heterozygotes (overdominance).
Both selections may increase heterozygosity up to the equi-
librium level, but only as a result of negative assortative
fertilization or selection in favor of heterozygotes (balancing
selection) can heterozygosity also exceed Hardy-Weinberg
expectations. This process can be detected with the aid of
isozyme gene markers if they are either involved in it di-
rectly or £ire in gametic linkage disequilibrium with selec-
tively significant genes (Crow and Kimura 1970).

In naturgJ conifer populations heterozygosity of adult
trees is often higher than in embryos (see Bush and Smouse
1992; Krutovskii and others 1988; Politov and others 1992,
for review). For instance, in Pinus radiata (Plessas and
Strauss 1986) heterozygosity of yoimg trees (17 to 20 years)
exceeded that of 5-year seedlings and in the latter it was
higher than in embryos. Yazdani and others (1985) found
higher heterozygosity of young £ind adult Scots pine trees
as compared to embryos; differences between the young and
the adult trees were slight indicating the action of selec-
tion exactly at early stages. For stone pines we can suggest
a pEirticular life stage of natural selection acting against
homozygotes — ^when groups of seedlings begin to germinate
out of nutcrackers' caches competing for resources. It is
noteworthy that Eurasian stone pines (as compared to white-
bark pine) display multistem forms rarely, which may be
the result of intensive competition between seedlings of the
same cache.

More direct evidence was obt£iined in experiments carried
out by Farris and Mitton (1984), who studied embryos and
seedlings of yellow pine {Pinus ponderosa) and foimd the
increase of heterozygosity with age. Muona and others
(1987) reveeded heterozygosity deficit in Scots pine embryos
and its "disappearance" in 1.5-year-old seedlings.

Inbred progeny elimination at early developmental stages
in conifers is also supported by some nonallozyme data (see
for example, Koski 1973).

CONCLUSIONS

Study of alloz5rme variability in stone pines has shown
the genetic diversity tjrpical to widespread and abimdant
conifer species with continuous ranges. Differences between
stone pines can be explained by both the events of their
recent evolutionary history and adaptive significance of
the allozyme variation. Heterozygote deficiency among em-
bryos and slight heterozygote excess in mature trees are
common features of stone pines, as well as other conifers.
A substantial part of self-pollinated progeny is eUminated
during ontogenesis by natural selection.

ACKNOWLEDGMENTS

We are greatly thankful to G. R. Furnier, V. N. Vorobjev,
N. A. Vorobjeva, L. I. Milyutin, R. T. Gut, and P. Raztmiov
for providing seeds and aid in their sampling, to Yu. S.
Belokon' for helpful assistance in the laboratory analysis,
to Yu. P. Altukhov for constant support of our study and
very useful conmients on this meinuscript, and to K. Ritland
for kindly providing us with the computer program MLT.

We also appreciate H.-R. Gregorius's invaluable comments
that helped to substantially improve the manuscript.

REFERENCES

Adams, W. T.; Strauss, S. H.; Copes, D. L.; Griffin, A. R.,
eds. 1992. Population genetics of forest trees: Proceed-
ings of the International Symposiimi on Population Ge-
netics of Forest Trees; 1990 July 31-August 2; Corvallis,
OR, U.S.A. Kluwer Academic Publishers. 420 p. (see also
New Forests. 1992. 6.)

Bobrov, E. G. 1978. Forest-forming conifers of the USSR.
Moscow: Nauka. 190 p. [In Russian].

Brown, A. H. D.; Barrett, S. C. H.; Moran, G. F. 1985. Mat-
ing system estimation in forest trees: Models, methods
and meanings. In: Gregorius, H.-R., ed. Population genet-
ics in forestry. Lecture Notes in Biomathematics. Vol. 60.
Berlin-Heidelberg-New York-Tokyo: Springer- Verlag:
32-49.

Bush, R. M.; Smouse, P. E. 1992. Evidence for the adaptive
significance of allozymes in forest trees. New Forests. 6:
179-196.

Critchfield, W. B.; Little, E. L. 1966. Geographic distribution
of the pines of the world. Misc. Publ. 991. Washington,
DC: U.S. Department of Agriculture. 99 p.

Crow, J. F.; Kimura, M. 1970. An introduction to popula-
tion genetics theory. New York: Harper and Row.

Farris, M. A.; Mitton, J. B. 1984. Population density, out-
crossing rate and heterozygote superiority in ponderosa
pine. Evolution. 38(5): 1151-1154.

Fineschi, S.; Cannata, F.; Malvolti, M. E.; Hattemer, H. H.,
eds. 1991. Biochemical markers in the population genet-
ics of forest trees: Proceedings of the Joint Meeting of the
Working Parties S2. 04-021 Population Genetics and Eco-
logical Genetics and S2. 04-05 Biochemical Genetics of
the International Union of Forest Research Organizations
(lUFRO); October 1988; Porano, Italy. SPB The Hague:
Academic PubHshing bv. 251 p.

Fumier, G. F.; Knowles, P.; Aleksiuk, M. A.; Dancik, B. P.

1986. Inheritance and linkage of allozymes in seed tissue
of whitebark pine. Canadian Journal of Genetics and
Cytology. 28(4): 601-604.

Fumier, G. R.; Knowles, P.; Clyde, M. A.; Dancik, B. P.

1987. Effect of avian seed dispersal on the genetic struc-
ture of whitebark pine popidations. Evolution. 41(3):
607-612.

Goncharenko, G. G.; Padutov, V. E.; Krutovskii, K V.;
Podzharova, Z. S.; Kirgizov, N. Ya.; Politov, D. V. 1988.
Level of genetic variability in Pinus sibirica in Altai.
Doklady Akademii Nauk SSSR. 299(1): 222-225. [In
Russian]. [English translation published by Pleniim
PubHshing Corp. in Doklady Biological Sciences; 1988
September: 139-141].

Hamrick, J. L.; Godt, M. J. 1989. Allozyme diversity in
plant species. In: Brown, A. H. D.; Clegg, M. T.; Kahler,
A. L.; Weir, B. S., eds. Plant population genetics breeding
and genetic resources. Sanderland, MA: Sinauer Associ-
ates: 43-63.

Koski, V. 1973. On self-fertilization, genetic load and sub-
sequent inbreeding in some conifers. Comm. Inst. For,
Fenn. 78(10): 1-40.

Krutovskii, K. V.; Politov, D. V.; Altukhov, Yu. P. 1987.
Genetic variability in Siberian stone pine, Pinus sibirica
Du Tour. I. (jenetic control of isozyme systems. Genetika.
23(12): 2216-2228. [In Russian]. [EngHsh translation
published by Plenimi Publishing Corp. in Soviet Genet-
ics; 1988 June: 1568-1576].

Krutovskii, K V.; Politov, D. V.; Altukhov, Yu. P. 1988.
Genetic variability in Siberian stone pine, Pinus sibirica
Du Tour. II. Level of allozyme variability in Western
Sayan. Genetika. 24(1): 118-124. [In Russian]. [EngHsh
translation published by Plenum Publishing Corp. in
Soviet Genetics; 1988 July: 81-86].

Krutovskii, K V.; Politov, D. V.; Altukhov, Yu. P.; Milyutin,
L. I.; Kxiznetsova, G. V.; Iroshnikov, A. I.; Vorob'ev, V. N.;
Vorob'eva, N. A. 1989. Genetic variability in Siberian
stone pine, Pinus sibirica Du Tour. IV. Genetic diversity
and differentiation between populations, (jenetika. 25(11):
2009-2032. [In Russian]. [EngHsh translation pubHshed
by Plenimi Publishing Corp. in Soviet Cjenetics; 1990 May:
1343-1362].

Krutovskn, K. V.; Politov, D. V.; Altukhov, Yu. P. 1990. In-
terspecific genetic differentiation of Eurasian stone pines
for isoenzyme loci. Genetika. 26(4): 694-707. [In Russian].
[English translation published by Plenum Publishing
Corp. in Soviet Genetics; 1990 October: 440-450].

Krutovskii, K V.; Politov, D. V. 1992. Study of intra- and
interspecific genetic differentiation between Eurasian
cedar pines using isozyme loci and multidimensional
analysis. In: Symposium on molecular mechanisms of
genetic processes: proceedings. Moscow: Nauka Publish-
ers: 87-96. [In Russian].

Lanner, R. M. 1990. Biology, taxonomy, evolution, and
geography of stone pines of the world. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposimn on whitebark pine ecosystems: ecology and
management of a high-mountain resource. Gen. Tech.
Rep. INT-270. U.S. Department of Agriculture, Forest
Service, Intermoimtain Research Station: 14-24.

Ledig, F. T. 1986. Heterozygosity, heterosis, and fitness
in outbreeding plants. In: Soule, M. E., ed. Conservation
biology: the science of scarcity and diversity: 77-104.

Mvdler-Starck, G.; Ziehe, M., eds. 1991. Genetic variation
in Eiiropean populations of forest trees: proceedings of
the workshop on genetic variation in forest tree popula-
tions in Europe; 1990; Gottingen, Germany. Frankfurt
am Main: Sauerlander's Verlag. 271 p.

Muona, O.; Yazdani, R.; Rudin, D. 1987. Genetic change
between life stages in Pinus syluestris: allozjone varia-
tion in seeds and planted seedlings. Silvae Genetica.
36(1): 39-42.

Plessas, M E.; Strauss, S. H. 1986. Allozyme differentiation
among populations, stands, and cohorts in Monterey Pine.
Canadian Journal of Forest Research. 16(6): 1155-1164.

Podogas, A. v.; Shurkhal, A. V.; Semerikov, V. L.; Rakitskaya,
T. A. 1991. Genetic variabiHty of needle tissue enzymes
of Siberian cedar pine (Pinus sibirica Du Tour). Genetika.

27(4): 695-703. [In Russian]. [English translation pub-
lished by Plenum Publishing Corp. in Soviet Genetics;
1991 October].
Politov, D. V. 1989. Allozyme polymorphism, genetic dif-
ferentiation, and mating system in Siberian stone pine
Pinus sibirica Du Tour. IVIoscow: Institute of General
Genetics, Russian Academy of Sciences. 190 p. Thesis.
[In Russian].

Pohtov, D. v.; Krutovskii, K. V.; Altukhov, Yu. P. 1989.
Genetic variability in Siberian stone pine, Pinus sibirica
Du Tour. III. Linkage relationships among isozyme loci.
Genetika. 25(9): 1606-1618. [In Russian]. [EngHsh trans-
lation published by Plenum Publishing Corp. in Soviet
Genetics; 1990 IVIarch: 1053-1063].

PoHtov, D. v.; Krutovskii, K. V. 1990. Genetic variability
in Siberian stone pine, Pinus sibirica Du Toiu*. V. Analy-
sis of mating system. Genetika. 26(11): 2002-2011. [In
Russian]. [English translation published by Plenmn Pub-
hshing Corp. in Soviet Genetics; 1991 May: 1309-1316].

Pohtov, D. v.; Krutovskii, K. V.; Altukhov, Yu. P. 1992.
Isozyme loci characteristics of gene banks of populations
of cedar pines. Genetika. 28(1): 93-114. [In Russian].
[English translation published by Plenvun Publishing
Corp. in Soviet Genetics; 1992 July: 76-95].

Ritland, K. 1990. A series of FORTRAN computer programs
for estimating plant mating system. Journal of Heredity.
81: 235-237.

Ritland, K.; El-Kassaby, Y. A. 1985. The nature of inbreed-
ing in a seed orchard of Douglas fir as shown by an effi-
cient multilocus model. Theoretical and Applied Genetics.
71: 375-384.

Ritland, K.; Jain, S. 1981. A model for the estimation of
outcrossing rate and gene frequencies using n indepen-
dent loci. Heredity. 47(1): 35-52.

Shaw, D. v.; Kahler, A. L.; Allard, R. W. 1981. A multilocus
estimator of mating system parameters in plant popula-
tions. Proceedings of the National Academy of Sciences
(USA). 78: 1298-1302.

Strauss, S. H.; Bosquet, J.; Hipkins, V. D.; Hong, Y.-P. 1992.
Biochemical and molecular genetic markers in biosys-
tematic studies of forest trees. New Forests. 6: 125-158.

Szmidt, A. E. 1982. Genetic variation in isolated popula-
tions of stone pine {Pinus cembra). Silva Fennica. 16(2):
196-200.

Szmidt, A. E.; IMuona, O. 1989. Linkage relationships of
allozyme loci in Pinus sylvestris. Hereditas. Ill: 91-97.

Tomback, D. F.; Linhart, Y. B. 1991. The evolution of bird-
dispersed pines. Evolutionary Biology. 4: 185-219.

Yazdani, R.; Lindgren, D.; Rudin, D. 1985. Gene dispersion
and selfing frequency in a seed-tree stand of Pinus sylves-
tris (L.). In: Gregorius, H. R., ed. Population genetics in
forestry. Lecture Notes in Biomathematics. Vol. 60. Berlin-
Heidelberg-New York-Tokyo: Springer- Verlag: 139-154.

GENETIC POPULATION STRUCTURE
AND GROWTH FORM DISTRIBUTION
IN BIRD-DISPERSED PINES

Diana F. Tomback
WiUiam S. F. Schuster

Abstract — ^At least nine species of the 20 wingless-seed pines
are dependent on the nutcrackers {Nucifraga) for seed dispersal.
Seed dispersal by nutcrackers influences several aspects of popu-
lation structure: (1) tree dispersion patterns and growth form dis-
tribution, (2) genetic relationships among trees within popula-
tions, and (3) genetic diversity among populations. Studies to
date of three species, whitebark pine {Pinus albicaulis), limber
pine {Pinus flexilis), and Swiss stone pine (Pinus cembra), pro-
vide examples of intrapopulation genetic relationships and growth
form distribution; other recent work provides some information
for interpopulation comparisons.

of parent trees, resulting in a haphazard distribution of
seeds with respect to seed source (Tomback 1978, 1982).
Also, nutcrackers often bury more than one seed per cache
(mean of three or fotir seeds per cache for Clark's nut-
cracker, Tomback and Linhart 1990), and the seeds may
come from the same parent tree (Tomback 1988; Tomback
and Knowles 1989).

In contrast, most wind-dispersed seeds fall within about
120 m of parent trees and are dependent on prevailing
winds for dispersal (see, for example, McCaughey and oth-
ers 1986). Updrafts and wind patterns associated with
storms may send wind-dependent seeds over longer dis-
tances and in other directions, and seed-storing rodents
may move seeds aroimd locally (Vander Wall 1992). How-
ever, because the seeds of these species usually move rela-
tively short distances and establishment is often in small,
disturbed patches, a family structure (local aggregations
of related genotypes) frequently results within popula-
tions (see, for example, Knowles 1984; Linhart 1989;
Linhart and others 1981; see also Fumier and others 1987;
Tomback and Linhart 1990).

These different seed-dispersal modes, bird vs. wind,
may lead to different genetic population structures. Ge-
netic population structure resiilts from nonrandom spatial
distribution of genotypes (Epperson 1990, and references
therein). Plant populations may show structure on differ-
ent scales, such as among populations and within popula-
tions. The common methods for investigating genetic
population structure are based on allozyme analysis,
comparing populations or subpopxilations by means of ei-
ther allele frequencies, Wright's or Nei's Gg.^, statistics,
or spatial autocorrelation.

Seed dispersal by nutcrackers influences several levels
of popiilation structure in wingless-seed pines. Within
populations, it affects both (1) tree dispersion patterns
(whether individuals are clumped vs. spaced) and, conse-
quently, growth form distribution and (2) genetic relation-
ships among trees. Among populations it affects genetic
diversity. In this paper, we explore the specific impacts
of nutcracker seed-dispersal behavior on each of these
aspects of population structure.

GROWTH FORM DISTRffiUTION

Of the 20 species of pines with large, wingless seeds
(collectively considered "stone pines"), nine are known
dependents on the nutcrackers, Nucifraga (Corvidae), for
seed dispersal (Tomback and Linhart 1990 and references
therein). Five of these bird-pines are in the Subsection
Cembrae, two in Strobi, and two in Parry a; for classifica-
tion see Critchfield and Little (1966). For several species
of these bird-dependent pines, other corvids, squirrels, or
other rodents may also make varying contributions to tree
estabHshment (see, for example, Hayashida 1989; Ligon
1978; Vander Wall 1992), but the factor in common to all
dispersers is the tendency to remove seeds from the vicin-
ity of parent trees and "scatterhoard" them (biuy small
clusters of seeds in subterranean caches for future use).
Clark's nutcrackers (A^. columbiana) bury one to 15 or
more seeds per cache at distances of several meters to
22 km from parent trees (Hutchins and Lanner 1982;
Tomback 1978, 1982; Vander Wall and Balda 1977). Eur-
asian nutcrackers (A^. caryocatactes) bury up to 24 seeds
per cache and travel similar distances (see, for example,
Bibikov 1948; Mattes 1982; Mezhenny 1961; Turcek
1966).

Thus, nutcrackers create a highly impredictable seed
shadow, with seedhng establishment oft^n occurring sev-
eral kilometers from parent trees (Tomback and others
1990). Nutcracker-dispersed seeds may end up in small,
isolated areas, particularly following disturbances such
as fires. In large, open or disturbed areas, different nut-
crackers may cache seeds from several difi"erent stands

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Diana F. Tomback is Associate Professor of Ecology, Department of
Biology and Center for Environmental Sciences, University of Colorado
at Denver, Denver, CO, 80217-3364. William S. F. Schuster is Forest Di-
rector, Black Rock Forest Consortium, Box 483, Cornwall, NY 12518.

Research to date on three species of bird-dispersed
pines — whitebark pine (Pinus albicaulis) (Fumier and
others 1987; Linhart and Tomback 1985), limber pine
(P. flexilis) (Carsey and Tomback 1992; Linhart and
Tomback 1985; Schuster and Mitton 1991), and Swiss
stone pine (P. cembra) (Tomback and others 1993) —

indicates that trees within populations occur in three
different growth forms. The single-trunk growth form
is typical of conifers and consists of a single genet (indi-
vidual). Also composed of a single genet, the multi-trunk
growth form has two or more trunks either contiguous
at the base or fused at the base or part way up the bole.
Multi-trunk trees may be the consequence of a tendency
toward side-branching and lack of apical dominance, par-
ticularly under stressful environmental conditions, such
as unstable soils, aridity, or mechanical damage to the
leader (Schuster and Mitton 1991; Weaver and Jacobs
1990). Although the tree cluster growth form is also
composed of two or more contiguous or fused trunks, the
trunks consist of different genets (individuals). This
growth form results from the germination and survival
of more than one seedling from a seed cache (growth form
terminology from Tomback and others 1990).

The latter two growth forms have similar morphologies
and thus cannot be distinguished with certainty without
genetic analysis (Schuster and Mitton 1991; Tomback and
others 1993) (figs. 1, 2). Consequently, in field surveys of
frequency distribution of different growth forms, we com-
bined the multi-trunk and tree cluster forms together in

Figure 1 — Swiss stone pine {Pinus cembra) tree
cluster. Protein electrophoresis indicated that each
trunk is a different genet.

Figure 2 — Swiss stone pine {Pinus cembra)
multi-trunk tree. Protein electrophoresis indi-
cated that both trunks had the same genotype.

the category "tree clumps" (term from Lanner 1980; the
same as "multi-trunk trees of unknown origin," Tomback
and others 1990). For reporting the results of field sur-
veys of tree clump frequency in populations, we use the
term "tree site" to refer to the small area of ground sup-
porting either a single-trunk tree, a multi-trimk tree, or
a tree cluster.

Overall, tree clumps accounted for 19 to 90 percent of
the tree sites surveyed in three species of bird-dispersed
pines (whitebark, limber, and Swiss stone) (table 1).
Whitebark pine in the two stands surveyed had the high-
est percentage of tree sites occupied by clumps. In the
two studies of limber pine combined, the percentage of
tree sites with clvmips ranged from 24 to 48; the latter
population occurred on a steep, southeast-facing slope
with gravelly soil at 3,310 m elevation. Carsey and
Tomback (1992) also examined 17 tree sites in a stand
of krummholz limber pine above timberline (3,460 m eleva-
tion) and found clumps at all tree sites. For Swiss stone
pine, clumps were found at 19 to 30 percent of the tree
sites and up to 45 percent of seedling sites. In the Piz da
Staz study area, the highest percentage of Swiss stone
pine clumps occurred at higher elevations, but this did
not hold for the Muottas Muragl study area.

Table 1 — Percentage of tree clumps (combined multi-trunk trees and tree clusters) in surveyed populations of three species
of bird-dispersed pines

Species

Maximum

Percent

and

Tree

trunks per

sites

location

Elevation

sites

clump

with clumps

References

Meters

Swiss stone pine

Switzerland

1,835

406

Tomback and

2,070

351

others 1 993

1 ,835

2,125

Whitebark pine

California

2,850

Tomback and

Montana

2,730

Linhart 1990

Limber pine

Colorado

1,650

361

Schuster and

Mitton 1991

Colorado

2,585

105

Carsey and

2,810

135

Tomback 1992

3,310

140

The preferred laboratory method for distinguishing tree stone pine clumps sampled in one stand, 70 percent were
clusters from multi-tnmk trees has been starch gel pro- tree clusters (Tomback and others 1993). Remarkably,

tein electrophoresis to assay the genotype of seed (Fiimier an examination of the number of gene loci used for each
and others 1987) or foliage tissue obtained from each study in table 2 shows no relationship with the percent-

tnmk of a tree clump (Carsey and Tomback 1992; Linhart age of clusters determined for each sample (Spearman
and Tomback 1985; Schuster and Mitton 1991; Tomback rank correlation, = -0.406, n = 12, NS). With the excep-

and others 1993; for review of technique, see Kephart tion of the Umber pine clumps analyzed by Schuster and

1990). Those clvunps that have two or more trimks with Mitton (1991), which represented all the climips in the

different genotypes are classified as tree clusters. Unfor- popialation, the clvmips examined in the studies in table 2

timately, this technique may underestimate slightly the
nimibers of tree clusters, because seeds within nutcracker
caches tend to be genetic relatives and thus may inherit
similar genotypes (Carsey and Tomback 1992; Schuster
and Mitton 1991). In a few cases tree cluster members
may possess identical genotypes at the gene loci exam-
ined, so the clusters are misclassified as multi-trunk
(single-genotype) trees. The proced\ires for calculating
the probability of such errors are presented in Schuster
and Mitton (1991).

For studies of whitebark pine, hmber pine, and Swiss
stone pine combined, genetic analysis indicated that 18
to 100 percent of the tree clumps sampled were tree clus-
ters (table 2). The low value, 18 percent, was obtained by
Schuster and Mitton (1991) for a limber pine population
in the Pawnee National Grasslands, at the periphery of
the range of both the pine and Clark's nutcracker. The
high values of 100 percent for limber pine and 83 percent
for whitebark pine were based on small sample sizes
in = six clumps in both cases, Linhart and Tomback 1985)
Filmier and others (1987) found that 58 and 70 percent,
respectively, of the whitebark pine clumps they sampled
in two stands contained more than one genotype; they did
not sample all trunks of some clumps, so these percent-
ages are conservative. Clumps in four limber pine stands
were studied by Carsey and Tomback (1992), with per-
centages of clusters ranging from 24 percent above tim-
berhne to 81 percent at 2,810 m elevation. Of the Swiss

Table 2 — Percentage of tree clumps (combined multi-trunk trees
and tree clusters) found to be tree clusters (growth form
with more than one genet) in three species of bird-
dispersed pines

Species

Clumps
examined

Gene
loci
used

Percent
clusters

References

Swiss stone

Tomback and others

pine

1993

Whitebark pine

Linhart and Tomback

1985

Furnier and others

1987

Limber pine

Linhart and Tomback

1985

100

108

Schuster and Mitton

1991

7-9

Carsey and Tomback

7-9

1992

7-9

represented a sample of the total number of climaps present
and were sometimes not selected entirely at random. For
the studies by Furnier and others (1987) and Carsey and
Tomback (1992), only reproductively mature tree clumps
were sampled; for the studies by Linhart and Tomback
(1985) and Tomback and others (1993), clumps were usu-
ally sampled as encountered, but an effort was made to
include some seedlings and young trees.

Only three studies permit an estimate of the overall
percentage occurrence of tree clusters among tree sites
in populations (table 3). Limber pine tree clusters in the
Pawnee National Grasslands population occurred at only
5 percent of the total tree sites. For limber pine in the
Front Range, clusters occurred at similar percentages of
tree sites, 19 to 21, despite differences in elevation. Clus-
ters of Swiss stone pine occupied 21 percent of the tree
sites in the upper subalpine population.

Apparently, tree clusters typically occupy a substantial
proportion of the tree sites in each population of whitebark,
limber, and Swiss stone pine. Moreover, the percentage
of individual genets in a population that are found in clus-
ters is much higher than the proportion of tree sites sup-
porting clusters. Based on the previous studies, 12 to 40
percent of the individuals in populations of bird-dispersed
pines occur in a highly clumped spatial pattern, primarily
as a consequence of seed dispersal by nutcrackers.

If stress plays a role in the growth of the multi-trunk
form from a single genet, as proposed by Schuster and
Mitton (1991) and Weaver and Jacobs (1990), then we
might predict a relatively higher percentage of multi-
trunk trees in more stressful environments. In fact, about
25 percent of the limber pine tree sites in the Pawnee
National Grasslands population support multi-trunk trees
(from tables 1 and 2). This area is at lower elevations
(1,650 m) on the plains east of the Rocky Mountains, at
the eastern boundary of the limber pine range, and trees
may be subjected to stress factors such as grazing, fire,
unstable slopes, heat, and extreme aridity (Lepper 1974;
McNaughton 1984; Schuster and Mitton 1991). The high-
est estimated percentages of multi-trunk limber pine trees
in the Front Range populations studied by Carsey and
Tomback (1992) occurred in the upper subalpine popula-
tion (27 percent, elevation 3,310 m) and in the population
above timberline (76 percent, elevation 3,460 m), where
winds and accompanying desiccation are extreme and the

Table 3 — Estimated percentage of tree sites tliat support clusters of
individuals in populations of two species of bird-dispersed
pines

Table 4 — Estimated mean and standard error of relatedness of
individuals within and between limber pine tree clusters

Percent

Species clusters Elevation

References

Meters

Swiss stone pine 21 2,050 to 2,240 Tomback and others 1 993

Limber pine

5
19
19
21

1,650
2,585
2,810
3,310

Schuster and Mitton 1 991
Carsey and Tomback 1 992

Mean Standard error

References

Within tree clusters

0.19 0.10 Schuster and Mitton 1 991

.43 .13 Carsey and Tomback 1 992

Between tree clusters

.01 .04 Carsey and Tomback 1 992

soil is gravelly and unstable. We also calculated that only
9 percent of the tree sites of the upper subalpine popula-
tion of Swiss stone pine supported multi-trunk trees. In
contrast to the former study areas, the Swiss stone pine
tree sites are not typically characterized by aridity and
poor soils, which may explain the relatively low frequency
of multi-trunk trees (Tomback 1988).

We should note here that tree clusters are not known
to occur in Colorado pinyon pine (Pinus edulis), which is
dispersed by both Clark's nutcrackers and pinyon jays
(Gymnorhinus cyanocephalus). Vander Wall and Balda
(1977) report that only one seed from each cache grows
to maturity.

GENETIC RELATIONSHIPS

Relationships Within and Between Tree
Clusters — Because nutcrackers tend to harvest and
pouch several seeds from the same pine cone and visit
more than one cone from the same tree or genet (if a clus-
ter) (Tomback 1988; Tomback and Knowles 1989), they
tend to place sibling seeds in the same seed cache. Dur-
ing a bumper whitebark pine cone crop (1989) in Yellow-
stone National Park, Tomback and Knowles (1989) ob-
served nutcrackers taking entire pouchloads of 28 to 97
seeds {x = 53 seeds, n = 8) from a single cone or individual
tree (genet) or trimk.

One consequence of this behavior is the tendency for
individuals in tree clusters to be genetic relatives — either
half-siblings, siblings, or the products of selflng. Furnier
and others (1987) found that individuals within tree clus-
ters of two populations were genetically more similar
(smaller genetic distance) than individuals in different
tree clusters. Individuals within clusters differed by a
mean of 1.66 alleles; individuals of different clusters dif-
fered by a mean of 5.82 alleles or greater (of 11 gene loci).
Two studies examined genetic relatedness within limber
pine clusters using the Quellar and Goodnight (1989) esti-
mator (table 4): For tree clusters in the Pawnee National
Grasslands population, Schuster and Mitton (1991) deter-
mined that individuals were related on average as half-
siblings. In addition, they determined that cluster mem-
bers that had fused together at the base were more closely
related on average (r = 0.35) than cluster members that
had not fused (r = 0.05). Carsey and Tomback (1992) cal-
culated that cluster members in their upper subalpine
population (3,310 m elevation) were related on average
as between half and full siblings; members of different
clusters were unrelated on average (table 4).

Possible Biological Implications of Tree
Clustering — Despite the fact that tree clusters occiir com-
monly in populations of bird-dispersed pines and include
a large proportion of the genets in the population, little is
known of the biology of this growth form (Tomback and
others 1993). Several papers have speculated about the
potential advantages and disadvantages to individuals
growing in clusters (Schuster and Mitton 1991; Tomback
and Linhart 1990; Tomback and others 1993).

Individuals in clusters may have higher survivorship
than single genets vmder certain environmental condi-
tions because of morphological or physiological benefits
(Bullock 1981; Keeley 1988; Mattes 1982). A sturdy, well-
anchored structure might better resist the strong winds
and mechanical damage of higher elevation environments.
Root grafting may also occur among cluster members, par-
ticularly relatives (Schuster and Mitton 1991), resulting
in efficient water and nutrient acquisition (Holtmeier
1986). In addition, since nutcrackers may establish pines
at some distance from conspecifics, for example, in burned
areas or disturbed patches, individuals in tree clusters
may cross-pollinate and thus avoid selfing (Tomback and
Linhart 1990). This might produce a high proportion of
sound seeds, particularly if clusters consisted of both rela-
tives and nonrelatives. Finally, both Schuster and Mitton
(1991) and Tomback and Linhart (1990) point out that the
occurrence of kin groups of bird-dispersed pines may re-
sult in kin selection as an evolutionary force. Possible
consequences for cluster members may include some de-
crease in competition or facilitation of rootgrafting.

With respect to the potential disadvantages to indivi-
duals growing in clusters, one area of possible fitness loss
is reproductive output. For limber pine trees in three
stands in the Colorado Front Range (same populations
studied by Carsey and Tomback 1992), cluster members
produced fewer male and female cones than did single-
tnmk trees during each of 3 years (Feldman 1991;
Feldman and Tomback 1991). Reduction of reproductive
output may result from competition among cluster mem-
bers for moisture, nutrients, and space. These effects may
be exacerbated for limber pine, which typically grows on
windy, xeric sites with poor soils in the Colorado Front
Range (Feldman 1991; Feldman and Tomback 1991).
Feldman (1991) and Feldman and Tomback (1991) noted
that trees in clusters had smaller diameters than did
similarly aged single-trunk trees, supporting the possibil-
ity of competition.

Another possible disadvantage to cluster members is
that their seeds may be more inbred than seeds produced
by single-genet trees, because of pollen exchange within
clusters (Furnier and others 1987; Tomback and Linhart
1990). PoUtov and Krutovskii (1990) provide some evi-
dence for this in Siberian stone pine (Pinus sibirica),
which is dispersed by the Eurasian nutcracker. Examin-
ing the mating system of Siberian stone pine in seven
populations, they noted an excess of homozygotes among
seed embryos in the majority of their samples. They at-
tributed this to a relatively high rate of self-pollination,
but pollination among related cluster members would
have a similar effect.

Genetic Spatial Structure Within Popvilations —

Several studies indicate that in many wind-dispersed
conifers there is family structure within popiilations
(local aggregations of genetic relatives) (see, for example,
Knowles 1984; Linhart 1989; Linhart and others 1981).
This is the consequence of seed dispersal typically re-
stricted to short distances (Ftunier and others 1987;
Tomback and Linhart 1990), and the availability of local,
disturbed patches for colonization (see, for example,
Linhart and others 1981). In contrast, Furnier and others
(1987) did not find any evidence of family structure
among the neighboring tree clusters in two whitebark
pine populations (although cluster members tended to
be related). They attributed this lack of family structiire
with respect to geographic distance to the seed-caching
behavior of Clark's nutcrackers. Since Clark's nutcrack-
ers do not defend caching territories, several different
nutcrackers may cache in one area, often bringing seeds
in from different stands of trees (Tomback 1978, 1982).
This probably results in a haphazard distribution of re-
lated seed caches (Furnier and others 1987; Tomback
and Linhart 1990). The Eurasian nutcracker in Sweden
(Swanberg 1956) and in the Alps (Mattes 1982) stores
many of its caches within territories defended by pairs.
However, the seeds cached in these territories are both
from local and more distant parent trees (up to 15 km.
Mattes 1982), so there is also considerable mixing in any
area. PoHtov and Krutovskii (1990) examined seven popu-
lations of Siberian stone pine and also foimd no evidence
for intrapopulation substructure.

GENETIC DIVERSITY

Generally speaking, conifers show less genetic diversity
among popiilations (mean = 0.068) than do other kinds
of plants (mean Gg^ = 0.250) (Hamrick and Godt 1990).
After an extensive review of the plant literature, Hamrick
and Godt (1990) concluded that animal-dispersed species
have more differentiated populations (higher Gg^ values)
than do wind-dispersed species. But, seed dispersal by
nutcrackers, which occurs routinely over large distances,
might actually result in higher levels of gene flow than
from seed dispersal by wind and thus lower levels of popu-
lation differentiation.

Few studies have examined interpopulation variation in
bird-dispersed pines. Schuster and others (1989) geneti-
cally analyzed two limber pine populations at extreme
ends of a 1,700-m elevational transect in the Colorado
Front Range. They determined that pollen flow along this
gradient was impeded by pollination phenology; popula-
tions separated by more than 400 m elevation did not
have overlapping pollination periods. For the populations
at each end of the transect, allele frequencies differed sig-
nificantly at eight of the 10 loci studied. However, the
calculated nimiber of migrants exchanged between the
populations per generation (A^^ = 11.1) was well above the
level required to overcome genetic drift. Given the very
different environmental conditions of the two populations,
Schuster and others (1989) concluded that natural selec-
tion rather than lack of gene flow may have differentiated
the populations. They also suggested that seed dispersal

by Clark's nutcrackers is likely to be "a more important
component of gene flow in limber pine than pollen trans-
fer, at least across elevational gradients."

Schuster and Mitton (1992) also compared five popula-
tions of limber pine across a 100-km range and found low
levels of differentiation (mean Gg^ = 0.035). Hamrick (un-
published data cited in Betancourt and others 1991) ex-
amined eight populations of pinyon pine (Pinus edulis)
and found a slightly greater differentiation (mean Gg^ =
0.077).

Nine populations of the Siberian stone pine were stud-
ied by Krutovskii £ind others (1989) to examine interpopu-
lation genetic differentiation. Their results indicated only
slight differences among populations. Using Nei's meas-
ures of diversity, they determined that on average about
98 percent of allozyme variation was attributed to intra-
population variability and only 2 percent to interpopula-
tion variability (mean Gg^ = 0.022). However, they did
find a significant correlation between geographical dis-
tance and genetic differences among populations, but only
for populations separated by 400 km or more. They also
found a higher level of differentiation among populations
within the same region than between regions, which may
reflect adaptation to local environmental conditions. The
high level of intrapopulational variation in Siberian stone
pine is very similar to what has been found for most wind-
dispersed conifer species (Krutovskii and others 1989).

EVOLUTIONARY IMPLICATIONS

Tomback and Linhart (1990) speculated that bird-
dispersed pines might be more prone to genetic drift than
are wind-dispersed species. The ranges of these pines,
particularly subalpine species, tend to be highly frag-
mented, and small, isolated populations are not uncom-
mon. These smaller populations may show founder ef-
fects, particularly loss of rare alleles. In addition, because
of the occurrence of tree clusters and the potential for
cross-pollination among cluster members, there may be
high rates of inbreeding in these populations. When effec-
tive population size is smaller than 100, genetic drift can
be far more potent in causing gene frequency changes
than natural selection (Futuyma 1986 and references
therein). It is possible that the evolution of cone, seed,
and tree morphology in bird-dispersed pines was in part
the consequence of new traits that arose from a combina-
tion of drift and inbreeding in small, isolated populations.
Nutcrackers, or ancestral corvids, selected for preferred
traits and spread these traits to other populations or
founded new populations with these traits.

If bird-dispersed pines have this population structure,
Tomback and Linhart (1990) make two predictions: "(1)
there is greater genetic variation among populations of
bird-dispersed pines than among wind-dispersed pines,
and (2) there are small, isolated populations of bird-
dispersed pines with highly differentiated gene pools com-
pared to one another and, particularly, compared to large
populations that may regularly experience some gene flow
from other large populations."

Although the earliest populations of bird-dispersed pines
may have had the characteristics described by Tomback

and Linhart (1990), the findings of Krutovskii and others
(1989), Schuster and Mitton (1992), and Hamrick (unpub-
lished data cited in Betancourt and others 1991) do not
support the first prediction, and it may well be that the
great distances over which nutcrackers disseminate seeds
provide effective gene flow. Betancourt and others (1991)
describe a disjunct population of Colorado pinyon pine in
northern Colorado that was the result of a founding event
probably about 400 years ago. Compared to other pinyon
pine populations, the average percentage of polymorphic
loci and nimiber of alleles per locus were reduced, although
levels of heterozygosity were not affected. These, how-
ever, are typical founder effects (Betancourt and others
1991; Futuyma 1986). Within the last century, the found-
ing population has given rise to four descendant popula-
tions, three with fewer than 50 trees each (Premoli and
others 1993). The two smallest, most distant descendant
populations had fewer alleles per polymorphic locus and
reductions in individual heterozygosity compared to the
larger populations. The mean Gg^ value of 0.060 indicated
little differentiation among populations. Because tree
clusters do not occur often in Colorado pinyon pine, these
new populations may not adequately test prediction 2.
However, they do demonstrate that seed dispersal by
birds results in small, isolated foimding populations.
Much more work is required to thoroughly characterize
the genetic population structure of bird-dispersed pines
and to address questions concerning the biology of tree
clusters.

ACKNOWLEDGMENTS

We thank Konstantin Krutovskii and Dmitri Politov for
providing translations of their papers £ind for interesting
discussions during the workshop. In addition, we are
grateful to Jeff Karron, Andrea Premoli, and Kathleen
Shea for helpful comments on an earlier draft of the pa-
per. Support for DFT for attending the SSPE Workshop
was provided by the College of Liberal Arts and Sciences,
University of Colorado at Denver.

REFERENCES

Betancourt, J. L.; Schuster, W. S.; Mitton, J. B.;
Anderson, R. S. 1991. Fossil and genetic history of
a pinyon pine (Pinus edulis) isolate. Ecology. 72(5):
1685-1697.

Bibikov, D. I. 1948. On the ecology of the nutcracker.
Trudy Pechorskogo-Ilychskogo Gosudarstvennogo
Zapovednika. 4(2): 89-112. [In Russian, translated
by L. Kelso].

Bullock, S. J. 1981. Aggregations of Prunus ilicifolia
(Rosaceae) during dispersal and its effect on survival
and growth. Madrono. 28: 94-95.

Carsey, K. S.; Tomback, D. F. 1992. Growth form distribu-
tion and genetic relationships in tree clusters of Pinus
flexilis, a bird-dispersed pine. Manuscript.

Critchfield, W. F.; Little, E. L., Jr. 1966. Geographic
distribution of the pines of the world. Misc. Publ. 991.
Washington, DC: U.S. Department of Agriculture, For-
est Service. 97 p.

Epperson, B. K 1990. Spatial patterns of genetic varia-
tion within plant populations. In: Brown, A. H. D.;
Clegg, M. T.; Kahler, A. L.; Weir, B. S., eds. Plant popu-
lation genetics, breeding, and genetic resources.
Sunderland, MA: Sinauer Associates: 229-253.

Feldman, R. 1991. Growth form and reproductive output
in limber pine: the cost of mutualism. Denver, CO: Uni-
versity of Colorado at Denver, Department of Biology.
46 p. Thesis.

Feldman, R.; Tomback, D. F. 1991. Growth form and re-
productive output in limber pine: the cost of mutualism.
Manuscript.

Fumier, G. R.; Knowles, P.; Clyde, M. A.; Dancik, B. P.
1987. Effects of avian seed dispersal on the genetic
population structure of whitebark pine populations.
Evolution. 4(3): 607-612.

Futuyma, D. J. 1986. Evolutionary biology. 2d ed.
Sunderland, MA: Sinauer Associates. 600 p.

Hamrick, J. L.; Godt, M. J. W. 1990. Allozyme diversity in
plant species. In: Brown, A. H. D.; Clegg, M. T.; Kahler,
A. L.; Weir, B. S., eds. Plant population genetics, breed-
ing, and genetic resoiirces. Simderland, MA: Sinauer
Associates: 43-63.

Hayashida, M. 1989. Seed dispersal by red squirrels and
subsequent establishment of Korean pine. Forest Ecol-
ogy and Management. 28: 115-129.

Holtmeier, F. -K. 1986. Uber Bamninseln (Kollektive)
an der klimatischen Waldgrenze — imter besonderer
Beriicksichtigung von Beobachtungen inverschiedenen
Hochgebirgen Nord-amerikas. Wetter und Leben. 38:
121-139.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Keeley, J. E. 1988. Population variation in root grafting
and a hypothesis. Oikos. 52(3): 364-366.

Kephart, S. R. 1990. Starch gel electrophoresis of plant
isozymes: a comparative analysis of techniques. Ameri-
can Journal of Botany. 77(5): 693-712.

Knowles, P. 1984. Genetic v£iriability among and within
closely spaced populations of lodgepole pine. Canadian
Journal of Genetics and Cytology. 26: 177-184.

Krutovskii, K. V.; Politov, D. V.; Altukhov, Y. P.; Milyutin,
L. I.; Kuznetsova, G. V.; Iroshnikov, A. I.; Vorob'ev, V. N.;
Vorob'eva, N. A. 1989. Genetic variability in the Sibe-
rian stone pine, Pinus sibirica Du Tour. IV. Genetic
diversity and the extent of genetic differentiation be-
tween populations. Genetika. 25(11): 2009-2032. [Trans-
lation published by Pleniun Publishing Corp. in 1991].

Laimer, R. M. 1980. Avian seed dispersal as a factor in
the ecology and evolution of limber and whitebark
pines. In: Dancik, B. P.; Higginbotham, K. 0., eds. Sixth
North Americsin forest biology workshop: proceedings.
Edmonton, AB: University of Alberta: 15-48.

Lepper, M. G. 1974. Pinus flexilis James and its environ-
mental relationships. Davis, CA: University of California
at Davis, Graduate Group in Ecology. 197 p. Dissertation.

Ligon, J. D. 1978. Reproductive interdependence of
pinon jays and pinon pines. Ecological Monographs.
48: 111-126.

Linhart, Y. B. 1989. Interactions between genetic and eco-
logical patchiness in forest trees and their dependent
species. In: Bock, J. E.; Linhart, Y. B., eds. Evolutionary
ecology of plants. Boulder, CO: Westview Press: 1-31.

Linhart, Y. B.; Mitton, J. B.; Sturgeon, K. B.; Davis, M. L.
1981. Genetic variation in space and time in a popula-
tion of ponderosa pine. Heredity. 46: 407-426.

Linhart, Y. B.; Tomback, D. F. 1985. Seed dispersal by
nutcrackers causes multi-trunk growth form in pines.
Oecologia. 67: 107-110.

Mattes, H. 1982. Die Lebengemeinschaft von Tannenhaher
imd Arve. Berichte Eidenossische Anstalt fur das
forstliche Versuchswesen, Birmensdorf. Nr. 241. 74 p.

McCaughey, W. W.; Schmidt, W. C; Shearer, R. C. 1986.
Seed-dispersal characteristics of conifers in the Inland
Mountain West. In: Shearer, R. C, comp. Proceedings —
conifer tree seed in the Inland Mountain West sympo-
sium; 1985 August 5-6; Missoula, MT. Gen. Tech. Rep.
INT-203. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermoimtain Research Station: 50-62.

McNaughton, G. M. 1984. Comparative water relations
of Pinus flexilis at high and low elevations in the cen-
tral Rocky Mountains. Laramie, WY: University of
Wyoming, Department of Botany. 59 p. Thesis.

Mezhenny, A. A. 1961. Food competitors, enemies and dis-
eases. In: Egorov, O. V., ed. Ecology and economics of
the Yakut squirrel. Moscow: Akademiya Nauk: 124-129.

Politov, D. v.; Krutovskii, K. V. 1990. Genetic variation of
Siberian stone pine Pinus sibirica Du Tom*. V. Analysis
of mating system. Genetika. 26(11): 2002-2011. [Trans-
lation published by Plenimi Publishing Corp. in 1991].

Premoh, A. C; Chischilly, S.; Mitton, J. B. 1993. Levels
of genetic variation captured by four descendant
populations of pinyon pine (Pinus edulis Engelm.).
Manuscript.

Quellar, D. C; Goodnight, K. F. 1989. Estimating relate d-
ness using genetic markers. Evolution. 43: 258-275.

Schuster, W. S. F.; Alles, D. L.; Mitton, J. B. 1989. Gene
flow in limber pine: evidence from pollination phenology
and genetic differentiation along an elevational transect.
American Journal of Botany. 76(9): 1395-1403.

Schuster, W. S. F.; Mitton, J. B. 1991. Relatedness within
clusters of a bird-dispersed pine and the potential for
kin interactions. Heredity. 67: 41-48.

Schuster, W. S. F.; Mitton, J. B. 1992. Gene flow in
limber pine. II. Data for five populations. Manuscript
in preparation.

Swanberg, P. O. 1956. Territory in the thick-billed nut-
cracker A^uci/ra^a caryocatactes. Ibis. 98: 412-419.

Tomback, D. F. 1978. Foraging strategies of Clark's nut-
cracker. The Living Bird. 16: 123-161.

Tomback, D. F. 1982. Dispersal of whitebark pine seeds
by Clark's nutcracker: a mutualism hypothesis. Journal
of Animal Ecology. 51: 451-467.

Tomback, D. F. 1988. Nutcracker-pine mutualisms: multi-
tnmk trees and seed size. In: Ouellet, H., ed. Acta XIX
Congressus Internationalis Ornithologici, Vol. 1; 1986
June 22-29. Ottawa, ON: University of Ottawa Press:
518-527.

Tomback, D. F. 1988. [Unpublished observations in the
Upper Engadine Valley of Switzerland, October 1988].

Tomback, D. F.; Knowles, J. W. 1989. Post-fire whitebark
pine seed dispersal by Clark's nutcracker in Yellowstone
National Park. Unpublished data and observations.

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird-
dispersed pines. Evolutionary Ecology. 4: 185-219.

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Coevo-
lution of whitebark pine and nutcrackers: implications
for forest regeneration. In: Schmidt, W. C; McDonald,
K J., comps. Proceedings — sjnnposium on whitebark
pine ecosystems: ecology and management of a high-
mountain resource; 1989 March 29-31; Bozeman, MT.
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department
of Agriculture, Forest Service, Intermountain Research
Station: 118-129.

Tomback, D. F.; Holtmeier, F. -K.; Mattes, H.; Carsey,
K. S.; Powell, M. L. 1993. Tree clusters and growth form

distribution in Pinus cembra, a bird-dispersed pine.

Arctic and Alpine Research. [In press].
Turcek, F. J. 1966. Uber das Wiederauffinden von im

Boden versteckten Samen durch Tannen-und

Eichelhaher. Waldhygiene. 6: 215-217.
Vander Wall, S. B. 1992. The role of animals in dispersing

a "wind-dispersed" pine. Ecology. 73: 614-621.
Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations

of Clark's nutcracker and the pinon pine for efficient

seed harvest and dispersal. Ecological Monographs.

47: 89-111.

Weaver, T.; Jacobs, J. 1990. Occvirence of multiple stems in
whitebark pine. In: Schmidt, W. C; McDonald, K J.,
comps. Proceedings — symposium whitebark pine ecosys-
tems: ecology and management of a high-mountain re-
source; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
156-159.

Basic Ecology

International Workshop
St. Moritz 1 992

EXPERIENCES WITH REPRODUCTION
OF CEMBRA PINE

Ernst Frehner
Walter Schonenberger

Abstract — ^As regards reproduction, cembra pine {Pinus cembra
L.) is a special case. If the greater part of the seeds are stolen by
nutcrackers, they should be harvested before they are ripe and
then matured. With correct treatment it is possible to store
seeds for up to 7 years. After soaking in water, germination be-
gan almost at once. The separate steps along the long process
from harvesting via seed extraction, maturation, storage, testing,
sowing, and transplantation to planting out are described in
detail.

in the Alps, and that the existing stands would diminish
considerably within a few generations.

Depending on the number of cones produced, the nut-
cracker stores seeds as winter provisions in nimierous
caches, usually numbering over 10,000, each containing
from three to five seeds, but only consimies about 80 per-
cent of its reserve. The remaining seeds contribute
greatly to the regeneration of cembra pine stands. The
nutcracker may transport seeds over a distance of 15 km
or more and over a difference in height of up to 600 m.

This study reports our experiences over many years of
seed £ind afforestation trials with cembra pine {Pinus
cembra L.), from seed harvesting to planting.

Cembra pine is a special case as regards reproduction
(Frehner and Fuerst 1992). The extreme conditions in its
natural environment at the upper timberline strongly in-
fluence flowering, fertilization, and seed maturation, so
that the quantity and quality of seed varies from year to
year (table 1). The results of repeated studies on cone
quality, maturation, and storage vary correspondingly.

The special features of reproduction in cembra pine be-
gin with seed collection. Except in full mast years, most
of the seeds have been eaten or collected by the nut-
cracker by the time of their natural maturation. Like the
seeds of many other tree and shrub species, those of
cembra pine require several months of dormancy. Matu-
ration of the seed and time of extraction is therefore very
important for the development of the embryo. To better
exploit fiill mast years, the durability of the seeds should
be extended through improved storage methods. It is also
important that stored seeds should be induced to germi-
nate and produce successful saplings at short notice. Also
in terms of sowing, transplantation, and potting, cembra
pine requires special conditions.

SYMBIOSIS WITH THE
NUTCRACKER

The importance of the nutcracker for the natural distri-
bution of cembra pine is undisputed. The detailed study
by Mattes (1982) on this relationship concludes that with-
out the nutcracker there would be no cembra pine forests

CONE COLLECTION

In years with normal, moderate fructification, the nut-
cracker begins to pick the seeds at medium altitudes as
soon as ripening has begun, about mid-August, and con-
tinues to do so until about the end of September. This
means that collecting cones is a race with the bird. If
cones are to be collected in years with partial mast, this
must be done at an early stage, while they are still un-
ripe, before the birds begin their robbery, and then stored
under optimvmi conditions for maturation.

In long-term observations, 19 percent of the years pro-
duced full mast, 55 percent partial mast, and 26 percent
failed to produce (Mattes 1982). As a rule it is only in fiill
mast years that enough seeds ripen on the tree to meike
harvesting worthwhile; in this case the nutcracker does
not have time to devour or carry away all the remaining
seeds.

Neither color nor size of cones has any influence on seed
quality or planting success. The costs of harvesting and
seed extraction are considerably greater for small cones.

With early harvesting it is very important that the
seeds should be allowed to ripen imder optimum condi-
tions of moisture and warmth, as the embryos are in a
very early stage of development (only up to 3 mm long).

Studies on protected cones attached to mother trees at
2,000 m have shown that the length of the embryo hardly
changes between mid-September and the end of October.
In contrast, embryos in cones harvested early (mid-
September) have elongated by 3-4 mm. This shows that

Table 1 — Characteristics of cembra pine seed crops

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Ernst Frehner is former Head of the Research Nursery, Walter
Schonenberger is Head of the Research Section Silviculture, Swiss
Federal Institute for Forest, Snow and Landscape Research, CH-8903
Birmensdorf, Switzerland.

Harvesting time per tree

Yield per tree

Number of cones per tree

Weight per cone

Seed yield (fresh weight)

Number of living seeds per cone

Thousand-seed weight

about 1 hour
about 15 kg cones
about 150-200
60-100 g
20-30 percent
50-70
300-400 g

Table 2 — Cone collection, maturation, and seed extraction

September Cone harvesting (embryo length up to 3 mm)

September Stratification of cones for maturation at 1 0-20 °C

for 8-1 0 weeks

November Seed extraction (embryo length 3-5 mm)

it is not worthwhile enclosing whole trees in nets or treat-
ing the cones with chemicals. Early harvesting and favor-
able conditions of stratification are preferable to late
harvesting.

CONE MATURATION

Stratification (storing cones in layers in a moist, airy
substrate) is the simplest and most certain technique of
accelerating ripening and reducing inhibition of germi-
nation so that the seeds are ready to germinate quickly.
Stratification simulates the natural conditions of moisture
and warmth prevailing in the forest, or may even be
better.

The fi-eshly harvested cones are stored in layers at a
proportion of one part cones to five parts substrate. Suit-
able substrates are washed sand or a mixture of sand and
peat. The substrate must be sieved so that its particles
are smaller than the seeds. The stratified cones should
then be stored in a cupboardlike container with good
drainage that protects the cones from attack by mice or
birds.

The development of the embryos depends greatly on the
temperature of the substrate and the water content of the
cones. Studies have shown that the embryos develop only
slowly at low temperatures, while high temperatures also
have a negative influence. The best results were obtained
with substrate temperatures alternating between 10 and
20 °C. It is advisable to keep the stratified cones in a pro-
tected place (for example, an unheated plastic tiinnel or
glasshouse), where the temperature may rise toward
25 °C during the day and sink to about 5 °C at night.
Naturally, the temperatiire within the substrate does not
vary so greatly.

Moisture is very important. The substrate should be
moist and airy, and the cones should be just moist enough
to prevent the seed coat from shrinking away from the
endosperm. Maturation to the point of seed extraction
proceeds much better under moist conditions than with
dry storage. With ripening under dry conditions the ger-
mination rate is 50 percent lower than that after moist
storage, while dormancy is 10 times longer.

SEED EXTRACTION

After 8 to 10 weeks' stratification, the seeds can be ex-
tracted. The cones, which are by this time beginning to
disintegrate, are sieved out of the substrate, left to dry for
a while, and then rubbed through finer sieves to separate
the seeds from scales and cone axes. The finer scales and
substrate particles can then be removed through further

sieving and washing in a water bath, either immediately
or after another ripening process. In a water bath the im-
fertilized seeds (empty seed cases) float to the top and can
be decanted. To summarize, the course of cone matura-
tion is shown in table 2.

SEED MATURATION

The treatment of the seeds after extraction depends on
their intended use and how long they are to be stored.

Under natural conditions it takes 2 to 3 years for the
seeds to germinate after the cones have fallen. This may
be due to many factors, for example, hard and imperme-
able seed cases, incomplete development of the embryo, or
inhibiting substances in the endosperm, in the seed case,
or elsewhere in the seed. In cembran pine, the main rea-
son is the low stage of development of the embryos: at the
time of harvesting the embryos are not more than 3 mm
long.

With artificial maturation using variable temperatures,
the period between harvesting and germination can be re-
duced to one winter. Here, the aim is to induce the em-
bryos to grow fi"om a length of 3 mm or less to the germi-
nation length of 7-9 mm (table 3). To promote embryo
growth even after seed extraction, the seeds are alter-
nately exposed to warmth and cold in a monthly rotation.
Exposure to warmth for periods of more than 4 to 6 weeks
does not positively influence embryo development. Great
attention should be paid to water content throughout the
whole period of seed treatment, as during cone storage.
The better the seed maturation, the better the embryo de-
velopment and, consequently, the better the seeds are
suited for drying and storage.

BASIC PRINCIPLES OF STORAGE

After drying, the seeds are placed in one of the con-
tainer types mentioned and stored imder cool conditions.
In the early stages they should be regularly inspected.

Water content:

• the lower the water content, the lower the storage
temperature;

• the lower the storage temperature, the longer the
seeds can be stored;

• the longer the seeds have been stored, the longer
they must be stratified before sowing.

Table 3 — Results of the maturation of cembran pine seeds
(harvested September 22) after warm-cold-warm
treatment

Features

September 23

February 28

Seed length (peeled)

9.5 mm

10.4 mm

Channel length

4.5 mm

8.5 mm

Embryo length

1.0 mm

8.4 mm

Seeds with visible embryos

11 percent

72 percent

Seeds ready to germinate

96 percent

76 percent

Storage containers:

• at high water content (>20 percent), in perforated
plastic bags;

• at low water content (<15 percent), in airtight con-
tainers accommodating 2-5 kg.

Storage rooms: climatized:

• for short-term storage with high water content,
1-3 °C;

• for fairly long storage with low water content,
-5 to -10 °C.

Freeze drying has proved unsuitable for cembran pine
seeds, because the water content is too high.

Fungal attack: for prevention, the seeds should be
treated with a fungicide and regularly examined as to
smell and condensed moisture indicating a too-high water
content in the containers; if this occurs, the seeds should
be dried immediately.

STORAGE OF MATURE SEEDS

Without treatment for reduction of water content the
seeds cannot be stored longer than one winter. With re-
duction of water content to about 25 percent, the seeds
can be stored at 1-3 °C for up to 3 years without great loss
in the germination rate.

Later trials showed that gently handled seeds could be
dried to a water content of 15 percent, and that under
storage at -5 to -10 °C the germination capacity did not
decrease for up to 7 years. A water content of more than
20 percent rendered storage below 0 °C iinfeasible.

THREE SEED TREATMENTS

After seed extraction three treatments for three differ-
ent storage periods can be defined as:

• Treatment 1: for immediate sowing without storage
(table 4).

• Treatment 2: for storage over 2 to 3 years (table 5).

• Treatment 3: for storage over 3 to 7 years (table 6).

GERMINATION TRIALS

Germination trials begin in December. The seeds are
sown in a sterile peat/sand in sterilized containers, being

Table 4 — Treatment 1 : for immediate sowing without storage

November Seed stratification outdoors for 5-6 months;

substrate three parts washed sand to one part
seeds

April Sowing (usually 30-50 percent of seeds lie-over)

Advantages: No need for climatization

Disadvantages: No storage possible, annual seed harvesting

necessary, high proportion of lie-over, high risk
of robbery by birds and mice

Table 5 — Treatment 2: for storage over 2-3 years with high water

content under refrigeration in plastic bags (for example, in
years with partial or full mast)

November Seed stratification for 4-6 weeks at 1 6-1 8 °C in

washed sand

December/January Sieving and cleaning of seeds; drying of seeds
for 1 -2 weeks at 1 6-1 8 °C to about
25 percent water content: embryo length
increases to 5-7 mm

January Storage of seeds in perforated plastic bags at

1 -3 °C over 2-3 years

October Seeds are soaked for 1 -2 days; stratification

outdoors for 6 months in washed sand

April/May Sowing (no lie-over to be expected)

Advantages: Allows storage for up to 3 years without lie-over

Disadvantages: Storage climatized at 1-3 °C necessary, high
water content, risk of rotting and fungal
infection, long stratification period

normally spread and pressed down, then covered with
about 1 cm sand or substrate. The trials are best con-
ducted in rooms without direct sunlight; the ideal tem-
perature is 16 to 18 °C. Alternating temperatures have
a positive effect. The seeds should be watered by hand
according to need. Storage in greenhouses with automatic
spraying has not proved suitable. The germination tests
take 6-8 weeks.

Table 6 — Treatment 3: for storage in airtight containers for 3-7 years
with low water content and freezing temperatures (for
example, for seeds harvested In good partial or full mast
years)

December

Seed stratification; first warm treatment at

16-18 °C for 4-5 weeks

January

Cold treatment at 1 -3 °C for 4-5 weeks

February

Second warm treatment at 1 6-1 8 °C for

4-5 weeks until germination begins; embryo

length increases to 7-9 mm

February

Sieving and cleaning of seeds; drying at

1 6-1 8 °C for 1 -2 weeks tol 5 percent water

content

March

Storage in airtight plastic bags or glass

containers at -5 to -1 0 °C for 3-7 years

April

Defrosting for 1-2 days, soaking in lukewarm

water for 1-2 days, sowing

Advantages:

Seeds can be stored for up to 7 years, no

lie-over, short-term stratification (6-8 weeks).

good mast years and seed quality can be

fully exploited

Disadvantages:

Long-term deep-freezing required

SOWING

The seeds are sown between the end of April and mid-
May. The number of seeds to be sown depends on the re-
sults of the germination trials. Because germination is
slow, the seeds should be sown in boxes resistant to mice
and birds. The best substrates are soils without lime-
stone mixed, according to conditions and pH, with peat
(which loosens the soil, maintains a pH between 5 and 6,
and regulates the moisture level) or raw humus contain-
ing mycorrhizal fungi from cembran pine stands. The
seed and transplantation beds should continuously be
occupied by cembran pines except for short periods with
green fertilizing.

Cembran pine grows best in thickly sown beds, with
1,200 to 1,800 seedlings per square meter, grown for 2
to 3 years in a seedbed.

If the seeds begin to germinate during stratification,
they should be sown immediately. If this is unfeasible,
they can be frozen at -3 to -4 °C and sown as required.
Seeds that have already germinated and whose radicle is
longer than 1 cm should not be planted in a seedbed, but
thinned out; otherwise the resulting plants are liable to
be worthless, with tortuous root collars, short stems,
small crowns, and few fine roots. Size, quality, and yield
are definitely better in thinned seedlings. Shortening the
roots causes deformations.

We did not run any trials with sowing in autumn, as
the risks of attack by mice and birds and also snow pres-
sure on the protection lattice are then very high.

TRANSPLANTING

The specially designed beds are continuously supple-
mented with acid soil (for example, peat) and if necessary
with chalk-free acid fertilizer. To hinder invasion by
weeds and frost heaving and to restrict evaporation, and

also to maintain a good soil structiire, the beds are
mulched.

It is recommended that 3-year-old saplings be trans-
planted in spring or summer (from mid-July). This can be
done by machine, with a spacing of 10 by 18 cm. Saplings
with poor root development should be discarded, as it is
not worth planting them out. The saplings stay in the
transplanting bed for 3 years.

POTTING

High-altitude afforestations are exposed to high risks
from extreme weather conditions, snow, and pests. Expe-
rience has shown that seedlings in peat-fiber pots succeed
better than those with naked roots, especially on diffic\ilt
sites, so potting is recommended. The production of pot-
ted plants has been described by Schonenberger and
others (1990). It is important to keep the pots well
watered. The saplings should only be planted out when
the root tips begin to penetrate through the peat-fiber
pots.

REFERENCES

Frehner, Ernst; Fiirst, Ernst. 1992. Vom Samen bis zur
Pflanze. Ein Erfahrungsbericht aus dem Forstgarten.
Ber. Eidgenoss. Forsch. anst. Wald Schnee Landsch.
333. 47 p.

Mattes, Hermann. 1982. Die Lebensgemeinschaft von
Tannenhaher und Arve. Eidg. Anst. forstl.
Versuchswes., Ber. 241, 74 S.

Schonenberger, Walter; Frey, Werner; Leuenberger,
Franz. 1990. Oekologie und Technik der Aufforstung
im Gebirge - Anregungen fiir die Praxis. Eidg. Anst.
forstl. Versuchswes., Ber. 325, 58 S.

Translation: Margaret Sieber

DISTRIBUTION AND ECOLOGY OF
SIBERIAN STONE PINE IN THE URALS

P. L. Gorchakovsky

Abstract — The northern and southern boundaries of the distri-
bution area of Siberian stone pine (Pinus sibirica) in the Urals
and in the adjacent plains are examined; the main sociological
factors limiting boundary positions are described. There is a
natural tendency of this species to spread northward and south-
ward, from places recently invaded, but this process is usually
stopped or turned back by human influence. Numerous occur-
rences of Siberian stone pine beyond its present continuous range
indicate that the former distribution area of this species was
larger in the past than at present.

Siberian stone pine (Pinus sibirica), or briefly Siberian
pine, is considered to be one of the most valuable species
of coniferous trees growing in Russia. Research on the
limits of natural distribution of Siberian pine, on their
causes and dynamics, is of indisputable theoretical and
practical interest. Research is especially necessary for
scientific substantiation of rational use of all forests, par-
ticularly those with Siberian pine, for the definition of
measures to protect it fi"om being exterminated, and for
solving the problems concerning possible cultivation of
this species in adjacent regions.

The general distribution area of Siberian pine includes:
(1) northeastern regions of the European part of Russia
(the farthest west locality is in the upper reaches of the
Vychegda River); (2) the Ural Mountain range; (3) west-
em Siberia; (4) middle and eastern Siberia (up to the
Aldan River); and (5) mountain ranges of Altai, the Trans-
baikal area, and northern Mongolia.

The main limiting factors of the distribution of this
species are: heat deficiency in the north; permafi*ost, peat-
Ismd, and severe winter in the northeast; water deficiency
and low air hvunidity in the south.

The western part of the area is located in the Ural
Mountains and in the northeastern regions of the Euro-
pean part of Russia adjacent to them. Siberian pine occu-
pies rather large areas within the territory of Sverdlovsk
and Perm provinces, Komi Republic, and within Knanty-
Mansy national district of Tyimien province.

THE NORTHERN LIMIT

The northern border of the distribution area (fig. 1)
passes along the Izhma-Pechora watershed, advances

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

P. L. Gorchakovsky is Professor, Doctor of Biology, Chief Scientist at
the Institute of Plant and Animal Ecology, Russian Academy of Sciences,
Ural Branch, 620219, Ekaterinburg, Russia.

considerably northward (up to 65°30' N. latitude) along
the valley of the Pechora River and even farther north-
ward (67° N. latitude — the northern polar circle). Then
it switches back south in the form of a tongue, to the foot-
hills of the elevated eastern part of the Russian plain.
From Telpos-Iz Mountain, 64° N. latitude, it leaps north-
ward along the eastern slope of the Poleir and Prepolar
Urals reaching 66° 10' N. latitude. It then recedes south-
ward in the space between the rivers, but returns north-
ward again in a tongue-shaped projection along the vsdley
of the Ob River.

Besides that, a few isolated occurrences of Siberian pine
are known outside its main distribution area.

The northern limit of Siberian pine does not reach the
extreme boundaries of the distribution of larch (Larix
sibirica) and spruce (Picea obovata). The well-known
Russian botanist B. N. Gorodkov (1929) supposed that the
advancement of Siberisin pine to the north is restricted be-
cause its seeds do not get ripe under the severe conditions
of the short growing season.

However, the latest observations demonstrate that
even at the northern limit of the distribution area seeds
of Siberian pine usually get ripe. The main reason likely
is that Siberian pine is not able to develop its root system
imder low soil temperature (Tyrtikov 1954). In cold soils,
the roots of Siberian pine ramificate comparatively weakly
and grow more slowly than the roots of spruce and larch.
Consequently, insufficient soil heating during the growing
season impedes normal development of root systems, and
this appears to be the main factor limiting the distribu-
tion of Siberian pine in the north. Under the severe cli-
matic conditions of northern Siberia, pine does not grow
on cold, peatland soils.

Close to the northern boundaries, in the upper reaches
of the Pechora River, Siberian pine selects the warmest
well-heated locations, inhabiting even the riverside lime-
stone outcrops and cliffs. Contrary to that, in the regions
farther south, for example in the middle and southern
Urals, Siberian pine never grows in such habitats because
the riverside limestone cliffs are too warm during daylight
periods.

In the plains adjacent to the Urals (the Pechora and
western Siberian lowlands), the far-north occurrences of
Siberian pine are restricted to the valleys of the large riv-
ers (the Pechora, the Ob, and some of their tributaries).
The bovmdary deviates northward in the form of narrow
strips along the river valleys. In marshy areas between
the rivers, however, where superficial bedding of perma-
fi'ost fi*equently occurs, the border line is farther south.
Under plain conditions, the river valleys characteristically
have more permeable ground, lower levels of permafi"ost
bedding, and warmer soils that are more favorable to

Figure 1 — Northern distribution limit of Siberian stone pine in tine Urals and in the adjacent
plains. * = isolated occurrences.

Siberian pine. Moreover, one should bear in mind that
the valleys of the large rivers that bring masses of
comparatively warm water from the south are character-
ized by a warmer local climate favorable to the growth of
Siberian pine.

At the Lyapin depression, in the region of Saranpaul,
Siberian pine- dominated forests are to be found on buried
palsa peatbogs covered by sandy loam deposits. On hill-
ocks, better conditions for drainage and heating of the su-
perficial layer of accumulated deposits favor the formation
of rather tall Siberian pine forests. On the other hand,
raised bogs with pools and ridges between the hillocks are
treeless. The level of permafrost here is rather high; ice-
lenses are foimd inside the peaty hillocks.

In the plains, at the extreme Hmit of its distribution
area, Siberian pine is usually represented by solitary
specimens dispersed within spruce-larch-birch forests;
trees usually are healthy, without signs of atrophy; their
tnmks reach 10 to 14 m in height.

Within the western and eastern outskirts of the Ural
Moimtain range, the Siberian pine border displaces north-
ward along meridionally oriented low foothills and moim-
tain ridges. There, this species grows on slopes or near
brooks and small rivulets imder moderate or slightly in-
creased (flowing) ground moisture conditions. The advance
northward is promoted by a variety of sites suitable for for-
est growth as well as by favorable drainage, better soil
heating, and consequently by lower depth of permafrost.

In the watershed of the Urals there are huge mountains
topped by woodless tundrahke "goltsy" summits, located

within the mountain tundra and high-mountain cold
desert belts (Gorchakovsky 1975, 1989). They are char-
acterized by a more severe climate, and consequently the
northern limit of Siberian pine is considerably displaced
southward.

Siberian pine almost absolutely escapes the moimtain
massif of the Prepolar Urals from the west; its distribu-
tion is restricted to the Pechora lowland; in rare cases it
occurs in the moimtain part along deep erosion-tectonic
valleys of the large rivers.

On the eastern slope of the Prepolar Urals, however,
Siberian pine ascends rather high in the mountains, al-
most to the upper forest limit. In the "subgoltsy" belt
(analogous to the subalpine belt), in the region of Gorodkov
Mountain (Sale-Ur-Oika), Siberian pine shows dwarfed
growth (2.5 to 5 m in height), exhibiting trunks thickened
at the basal portion of the tree and tapering upward, and
with wide-branched crowns of low stability. Frequently,
such dwarfed trees bear abundant cones. At lower eleva-
tions, within the mountain-taiga belt in the plain, Sibe-
rian pine reaches 10 m in height. From the eastern side,
along the foothills, this plant extends to the Polar Urals
reaching 66° 10' N. latitude in the mountains of the so-
called Small Urals.

More intensive advancement of Siberian pine north-
ward, along the eastern slope of the Ural Mountain range,
was favored partly by historical causes. This species in-
vaded from Siberia to new places in the Urals. It arrived
earher on the eastern slope of the Ural Mountain range,
favored mainly by the specific natural conditions of the
eastern slope. The eastern slope of the Prepolar and

Polar Urals has a more continental climate than the west-
ern slope. Relative air humidity is lower here; annual
precipitation including that in the winter is less heavy;
and snow cover melts earlier in spring. Consequently, the
growing season is longer, and the top soil layer is better
heated.

On its northern boundaries, Siberian pine is distributed
disjimctively and occurs as dispersed small clumps or soli-
tary specimens in the taiga. All these extreme localities
are rather disjimct from each other. This may be ex-
plained primarily by the fact that only in rare locations
are the combinations of various environmental factors
favorable for growth of Siberian pine at the limits of its
range. On the other hand, the dispersion of this species
has been reduced here and there due to forest fires, exten-
sion of bogs, trampling of seedlings and young growth
by reindeer, and extermination by humans. In locations
where human activity is reduced to a minimum, Siberian
pine invades new places north of its present limit. This
advance is observed especially clearly in the mountain
regions of the Urals.

THE SOUTHERN LIMIT

The southern borderline of continuous distribution of
Siberian pine (fig. 2) passes through the Russian plain
near the city of Perm; then it turns southward along the
western slope of the middle Urals, reaching the most
southern position at 56° 10' N. latitude; then it switches
back north following the eastern slope of the middle Urals
until retiring into western Siberia.

A few outliers occur beyond this limit. The most inter-
esting of them are to be found in the mountain dark conif-
erous taiga of the southern Urals in the region of the town
of Zlatoust and in the Ilmen Mountains.

At the southern limit of its range, Siberian pine usually
grows in the form of single-standing, tall trees. Only on
rocky outcrops does it exhibit prostrated growth forms
(krummholz).

Natiiral regeneration of Siberian pine near the southern
boimdary of its distribution is generally satisfactory. How-
ever, there are many factors unfavorable to regeneration,
such as seed collection difficulties, fires, and trsimpling of

Figure 2— Southern distribution limit of Siberian stone pine in the Urals
and in the adjacent plains. * = isolated occurrences.

seedlings by livestock. In locations where such vmfavor-
able influences are absent, or they manifest themselves
to a lesser degree, regeneration of Siberian pine is usually
good. So, for example, Siberian pine reproduces very suc-
cessfully on raised sites protected from fire within the
Bakhmet Bog, which is surroimded by Scots pine forests
in the basin of the Pyshma River.

Along the western slope of the Urals, as well as in cer-
tain sites within the watershed zone where the climate is
more humid, Siberian pine has invaded new places signifi-
cantly farther south than in the Preural and Transural re-
gions. Its southern limit projects southward in the form
of tongues that comprise the western slope and part of the
watersheds of the moimtain range. Climate of the eastern
slope of the middle Urals is drier; the amoiint of precipita-
tion is less compared to the western slope. Therefore, on
the eastern slope of the movmtain range and in the foot-
hills zone adjacent to it (the region of predominance of
Scots pine forests), Siberian pine occurs considerably far-
ther north.

Within the peatland between the rivers of the western
Siberian plain, the southern limit of Siberian pine turns
south again. There Siberian pine grows mainly on raised
sites within bogs, where it is well protected from fire.

Siberian pine invaded the southern Urals later than
spruce and fir {Abies sibirica) and has not had enough
time in the prehistoric period to reach the climatically
influenced limit of its distribution. Human activities
hampered the southward advance of Siberian pine into
new places along the Ural Mountain range. Here and
there, numbers of this species have declined. Several cen-
turies ago Siberian pine had been distributed farther
south in the Urals compared to its contemporary limit.
Since then it has receded northward due to forest fires
and himian impact. There is clear evidence of recent de-
cline in total distribution. In the southern part of the
area, Siberian pine occurs sporadically in small groves,
groups of trees, or as single trees. It grows in places least
endangered by fire, most frequently on moist soils in the
headwater regions and valleys of brooks and rivulets, on
the margins of bogs and on raised sites within them, on
lake shores (somewhere close to water), and on steep out-
crops of acid rocks.

Formerly, the region of continuous distribution of Sibe-
rian pine extended south at least to the isolated localities
mentioned earlier, but probably this tree species extended
even farther south. Literature data as well as oral state-
ments by local residents provide evidence of it. Siberian
pine disappeared from these places not only due to cutting
of timber and forest fires, but also because local inhabit-
ants picked all the cones, often before cones were com-
pletely ripe. They also dug up seedlings for transplanting
into personal gardens.

DISCUSSION AND CONCLUSIONS

The rather complicated picture of distribution of Sibe-
rian pine in the Urals depends to a large extent on cli-
matic difl'erences between the relatively himiid western
slope of the range and the more continental eastern slope.

Humidity of the western slope restricted the northward
advance of Siberian pine, but favored its advance to the
south. On the contrary, continentality of the eastern
slope favored the advance of this species northward, but
restricted its advance southward.

At both the northern and southern Umits of its distribu-
tion, Siberian pine does not occur high up in the moun-
tains. Its northernmost occurrences in high mountains
are in the eastern part of the Prepolar Urals.

In the northern Urals, Siberian pine is widespread al-
most everywhere in the high-mountain belts. On certain
mountains (Oika-Nyor, Pas-Nyor, Chistop, and others),
however, it forms the upper forest limit. High-mountain
Siberian pine-dominated forests grow on steep slopes with
poorly developed soils. Near the upper forest limit, such
forests may be foimd mainly in less high mountains far
removed from large mountain massives. In such isolated
mountains, the upper forest limit is caused chiefly by
edaphic factors, while the cHmate of their treeless zone
and of the subgoltsy belt is less severe. Dwarfed, pros-
trate, and, in most cases, noncone-producing specimens
of Siberian pine of the northern Urals occur in high-
moimtain tundras significantly higher in elevation than
the normal upper forest limit. For instance, dwarf speci-
mens occur on the movmtains Isherim, Oika-Chahl,
Yalping-Nyor, Denezhkin Kamen, Konzhakovsky Kamen,
Kosvinsky Kamen, and others. The occurrences of kmmm-
holz growth forms above the normal upper forest limit
have to be explained by the dispersal of seeds into the
moimtain tundra zone by the nutcracker and rodents.

In the middle Urals near the simamits of relatively high
moimtains, Siberian pine also forms singly standing,
dwarfed trees (the mountains Kachkanar, Starick-Kamen,
and others). South of Moimt Starick-Kamen, this species
is never foimd in high-movmtain belts, but in the forest
belt its area spreads considerably farther south.

Judging by its biological and ecological features, Sibe-
rian pine may not be considered as a species becoming
extinct. It is perfectly well adapted to the taiga environ-
ment; it is a good cone producer, and, if human impact
does not interfere, it regenerates quite satisfactorily. Al-
most everywhere, regeneration occurs close to old trees.
However, one should bear in mind that Siberian pine
grows more slowly than other shade-tolerant coniferous
trees. Therefore, ground fires that annihilate second-
growth trees adversely afl'ect natural regeneration of this
species.

The border lines of Siberian pine, especially on the
southern border, have been reduced by careless and some-
times barbarian treatment of this valuable plant. It is
necessary to secure proper protective measiires for its
southern localities in the Urals and to use seeds gathered
from trees mainly for sowing under the forest canopy and
in tree nurseries.

In the coTorse of intensive harvest cuttings in the dark
coniferous fir-spruce forests that have individual Siberian
pine trees in the admixture, foresters usually retain Sibe-
rian pine as seed trees for natural regeneration. How-
ever, in many cases, this does not work well because in
the heavy clay and loamy soils of the plains, and in the

mountains on poorly developed soils, such trees r\in the
danger of windfall. To avoid windfall, it is necessary to
retain clumps of Siberian pine with accompanying tree
species instead of single specimens of Siberian pine.

Natural conditions of the middle Urals and of the
moiintain-range portion of the southern Urals (at least
up to 54° N. latitude) are favorable for regeneration and
growth of Siberian pine. This valuable tree species may
be reintroduced to areas south of its present distribution.
This can be concluded from the successful cultivation of
Siberian pine south of its present border line in the moun-
tain taiga regions of the middle and southern Urals.

REFERENCES

Gorchakovsky, P. L. 1975. The plant world of the high-
mountain Urals. Moscow: "Nauka" Publishing. 284 p.
[In Russian].

Gorchakovsky, P. L. 1989. Horizontal and altitudinal
differentiation of the vegetational cover of the Ural
Mountains. Pirineos, N 133: 33-54.

Gorodkow, B. N. 1929. The Polar Urals in the upper
reaches of the rivers Voykar, Synya and Lyapin. In:
Materials of the Committee for Expeditional Re-
searches, V. 7, Leningrad: 1-28. [In Russian].

Tyrtikov, A. P. Growth of tree roots at the northern limit
of forests. Bull, of Moscow Soc. of Naturalists. Section
of Biology. 59(1): 72-82. [In Russian].

ECOPHYSIOLOGICAL INVESTIGATIONS
ON CEMBRAN PINE AT TIMBERLINE IN
THE ALPS, AN OVERVIEW

Rudolf Hasler

Abstract — During the nineteenth century foresters as well as
scientists became interested in learning more about cembran
pine (Pinus cembra) in the Alps, but extensive ecophysiological
fieldwork did not begin before the 1930's. Great progress was
made after the estabUshment of permanent field stations at the
timberUne at Obergurgl (1953), Patscherkofel near Innsbruck
(1963), and Stillberg near Davos (1959). In a brief overview,
some selected results of ecophysiological research in cembran
pine are presented.

The old and bizarre cembran pines (Pinus cembra L.),
which form the timbedine over a wide range in the Cen-
tral Alps, have always fascinated himaans. In earlier
days, the tree was of some importance. It produced not
only wood but also nuts that could be used as food. There-
fore, this tree has been of some interest for a long time,
and a lot of books and papers about cembran pine in the
Alps have been published, in forestry as well as in biologi-
cal science.

This short overview presents only some selected ex-
amples of the ecophysiological research, with a special
focus on gas exchange. A general overview on ecophysio-
logical problems at the timberline was published by
TranquilUni (1979). Many other subjects, such as growth,
genetics, and mycorrhiza, have been investigated in
Austria, France, Italy, and Switzerland but are not men-
tioned in this overview.

BEGINNING OF INVESTIGATIONS

In the middle of the nineteenth century, foresters be-
came interested in the cembran pines growing in the Alps.
Fankhauser (1853) described the distribution of these
trees together with the climatological conditions they
need for growing well. He mentioned that it is not
possible to transfer this tree species to low elevations.
Kemer (1866) and Simony (1870) pubhshed similar data
for parts of Austria. At the same time, the diseases and
the insect enemies of the cembran pine in the Alps at-
tained some importance. They have been described by
Keller (1890) and others.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Rudolf Hasler is Head of the Section "Forest and Climate," Swiss Fed-
eral Institute for Forest, Snow and Landscape Research, CH-8903
Birmensdorf, Switzerland.

ECOLOGICAL AND GEOGRAPfflCAL
OBSERVATIONS

More and more scientists became interested in cembran
pine at the end of the nineteenth century and the early
twentieth century to learn more about these trees, which
grow on the most extreme sites in the mountains. Mono-
graphs concerning cembran pine in Switzerland (RikH
1909) and in Austria (Nevole 1914) were pubhshed. Many
other publications were written at this time. The main
subjects were growth, and again fungal diseases and prob-
lems with insects (Fankhauser 1903; Keller 1901, 1910;
and others).

FIRST ECOPHYSIOLOGICAL FIELD
EXPERIMENTS

Ecophysiological investigation in the field began in the
thirties on Patscherkofel near Innsbruck (Austria). To
work near the timberline in the Alps was a logistic prob-
lem. The cembran pines only grow near the timberline,
at high altitudes. It was extremely hard to work there in
earher times.

Cartellieri (1935) was probably the first to measiu*e
photosynthesis and water relations throughout a whole
season in cembran pine at timberline. At the same time
he also took measurements of different dwarf shrubs at
high elevation. He measured photosynthesis by absorbing
the carbon dioxide in barite-water, and then he deter-
mined the COg assimilation by titration in the laboratory.
Some idea about the difficulties of working at the timber-
line in those days is given in his papers: missing data in
the annual course are explained as "walking back to the
laboratory I lost some of the probes because some of the
glass flasks were broken," or in another example, "at this
time I got ill and had to stop the investigation." Never-
theless his results show more or less the same values as
we are measuring today with ovu* modem climatized and
very expensive instrumentation. Also about the same
time Ulmer (1937) investigated the annual course of fi*ost
hardiness and measured in parallel the osmotic values
in the needles.

PERMANENT FIELD STATIONS AND
CONTINUOUS CO2 MEASUREMENTS

Great progress in timberline research in the Alps was
made in the 1950's. Again in Austria, a permanent
experimental field station at an altitude of about 1,950 m
above sea level (a.s.l.) was established in 1953 near

Obergurgl. At the same time, the newly developed infra-
red gas analyzer (IRGA) made it possible to measm'e
continuously carbon dioxide in the air.

Tranquillini (1955) conducted experiments at the new
field station at Obergurgl to investigate the relations be-
tween light, temperature, and photosynthesis in needles
of young cembran pines growing in the sun or in the shad-
ows. He determined the light compensation points for
these needles as 500 and 150 Lux, respectively, and the
light saturation points as 25 and 18 kLux, respectively.
The COg assimilation values he reported reached 2.5 mg
and 2.1 mg COJg dry weight, respectively. Net photo-
synthesis starts at -5 °C and has an optimum range be-
tween 10 and 15 °C. The upper temperature limit is at
35 °C.

The microclimatic conditions during wintertime and
needle and soil temperatures as well as gas exchange and
water relations in young cembran pines growing on natu-
ral sites at the timberline were published by Tranquillini
(1957). He showed that in wintertime, even in warmer
weather conditions, no net photosynthesis was detectable.
Holzer (1958) explained this lack of photosynthesis by a
changed cell structure as well as by a changed physiolo-
gical behavior during wintertime.

Pisek and Winkler (1958) measured the annual course
of the capacity of net photosynthesis and of respiration in
detached shoots of cembran pines growing at different al-
titudes. During wintertime, they found an always nega-
tive carbon dioxide balance in the trees at the timberline,
whereas those growing at lower elevations could produce
a positive one, at least during periods with temperatures
above the freezing point. Keller (1970) confirmed these
findings, using another method. Tranquillini (1958,
1963d) investigated the annual course of frost resistance.
He found that young trees are more sensitive to frost than
mature trees. Young trees become damaged by tempera-
tures between -15 to -30 °C, while mature trees can sur-
vive temperatures lower than —45 °C. Keller and Beda-
Puta (1973) analyzed stem respiration in cembran pine
during wintertime.

Pisek and Winkler (1959) studied photosynthesis in re-
lation to temperature in different light conditions as well
as at different altitudes. With increasing light intensity,
the temperature optimum is shifted to higher tempera-
tures. There are some difficulties in comparing all these
older results with the measurements of today, because not
all parameters can be transformed (for example, illumina-
tion [Lux]). Some other values, such as the needle sur-
faces, were normally not used as a reference for gas ex-
change measurements before the 1980's.

A COg budget for a whole year was estimated by
TranquiUini (1959a, 1959b, 1963a, 1963b). Total CO^
uptake per year (gross photosynthesis) of a 5- to 8-year-
old tree at 2,000 m a.s.l. was about 7.8 kg. About 38 per-
cent was lost by respiration from needles, branches, stem,
and roots. According to the measured net COg assimi-
lation, a tree should produce 2.2 g dry matter per gram of
needles. In reality they produced 0.65 g. The author
mentioned that most of the missing carbon dioxide may
have been used by the mycorrhiza.

In the second half of the 1950's, ecological investiga-
tions with the aim of finding methods for successful

afforestation at high altitude were started in Switzerland
(Stillberg, Davos: 2,000-2,230 m a.s.l.) and in Austria
(Haggen, Sellraintal: 1,730-1,900 m a.s.l.). Cembran pine
was one of the species used in these field experiments,
which still are going on. It takes a long time to get results
from trees growing at high elevation near the timberline.
Svmimaries containing prelimin£iry results were published
by Schonenberger and Frey (1988) for Stillberg and by
Kronfuss (1980, 1983, 1986) for Haggen.

LABORATORY CLIMATE CONTROL

The technical possibilities developed fast. On
Patscherkofel (1,950 m a.s.l.) a climate laboratory was
established in 1963. In this laboratory a device was in-
stalled that allows analysis of the gas exchange of yoimg
trees in climatized windchannels at high altitude. Now it
became possible to control exactly the climatic parameters
such as temperature and air and soil humidity as well as
windspeed, which was a brand-new innovation. Using
this new instrumentation old measurements could be ve-
rified and a lot of new possibilities were opened up. In
particular, water relations could now be investigated un-
der controlled climatic conditions.

Tranquillini (1963c) investigated the CO^ assimilation
in relation to air and soil humidity. In water-saturated
soil, dry air (25 percent relative hmnidity) reduced photo-
synthesis in cembran pines by about 33 percent. A little
less water in the soil (half of field capacity) allowed some
higher photosjnithetic rates. Resistance to dry himiidity
increases from spring to autimin (Tranquillini 1965).

Pisek and others (1967, 1968, 1969) explored the tem-
perature limits of photosynthesis to chill and frost, and to
high temperatures. COg assimilation is operative between
-4.7 and +36 °C, but the trees can survive temperatures
between below -50 and up to 48 °C. Of great importance
is not only the time factor, but also the physiological con-
dition of the plant. Temperature hardiness changes with
the seasons, £md young trees or seedlings are more sensi-
tive to extreme temperatures than mature trees. The re-
lation between photosynthetical capacity and temperature
was analyzed by Tranquillini and Machl-Ebner (1971).
Another factor investigated was day length. Bamberg
and others (1967) and Schwarz (1970a, 1970b, 1971) have
shown that day length does not influence the photosyn-
thetic capacity during winter depression, but trees kept
in warm conditions during winter commenced growth in
spring much earlier than plants in nature. The variation
in frost resistance of the trees changed in parallel with
the temperature. In long-day regimes, frost resistance of
cembran pine remained at a lower level than those grow-
ing imder short-day conditions. The researchers supposed
that an endogenous rhythm also influences resistance to
extreme temperatures.

Caldwell (1970), using the climatized wind tunnel of the
Patscherkofel laboratory, has shown the effect of wind on
stomatal aperture, photosynthesis, and transpiration.
He did not detect a great influence of low windspeeds on
these parameters. At higher windspeeds (up to 8 m/sec)
photosynthesis and transpiration were slightly reduced.
Baig and others (1974) and Baig and Tranquillini (1980)
investigated the cuticular transpiration rate of cembran

pine and found that cuticular transpiration in this tree
is very slow but, as in other species, cuticular resistance
decreases in paredlel to the elevation at which the shoots
were collected.

Havranek (1972) found that cold soil temperatures
reduce photosynthesis and transpiration in cembran pine.
Havranek and Benecke (1978) investigated the influence
of soil moistiire on transpiration and photosynthesis.
Cembran pine begins to close the stomata at a soil water
potential of about -0.4 bars. Compared to other species,
cembran pine maintains a high water-use efficiency and
uses limited soil water economically. Giinthardt and
Wanner (1982) showed that needles from timberline trees
have on the average more wax than those grown at a
lower elevation.

Another step to get more precise resvilts for the CO^
balance of cembran pine was taken by Tranquillini and
Schiitz (1970). They measured bark respiration at the
timberline and reported that b£irk respiration also shows
an annual course. In general, bark respiration of cem-
bran pine was rather high compared to that of other spe-
cies. The loss of carbon dioxide by stem respiration
throughout a year was calculated to be about 18.5 percent
of gross photosynthesis. Measuring stem respiration con-
tinuously, Havranek (1981) found that there is a marked
fluctuation in stem respiration during the growing season,
but he could not find any relation between activity of stem
respiration and tree ring growth.

Net photosynthesis

40-

■a 32"
E 2411

o 16
E 81
E

-8

I I I I I B I I I I I 11 I I I I I 11 I I I I I 11 I I I I I 11 I I I I I I I I I I I 11 I I I I I H I I I I I 11 I I I

5 15 2515 15 2515 15 2515 15 25 15 15 25 1 5 15 251 5 15 2515 15 2515 15 25 I 5 15
Photosynthetically active radiation

'l I I I I I 11 I I I I I 11 r r I I I 11 I I 'V r i iTT i i i i ii T i i i i i i i i i i-ir i i i i i n i i i i i ii i i r
15 15 2515 15 2515 15 251 5 15 25 I 5 15 25 I 5 15 2515 15 25 I 5 15 2515 15 25 I 5 15

Air temperature

5 15 2515 15 2515 15 251 5 15 25 I 5 15 25 1 5 15 2515 15 2515 15 251 5 15 25 I 5 15

Soil temperature

I M III I I I I I I I I I I I I I I I I I I 11 I I I I I 11 I I I I I I I ( I I I 11 I I I I I 11 I I I I I 11 I I !'

1 5 15 2515 15 25 1 5 15 2515 15 25 1 5 15 2515 15 251 5 15 25 I 5 15 2515 15 2515 15

Water vapor pressure difference

Is 1525'!

15 15 2515 15STr'^l525

Sep Oct

Nov

5 15 25 15 15 2515 15 251

Dec Jan Feb Mar

5 15

Apr May Jun

25
20
15
10
5
0
-5
-10
-15
-20

Figure 1 — Photosynthesis in cembran pine,
photosynthetically active radiation, air and soil
termperature, and air humidity measured at the
timberline (Stillberg, 1,980 m a.s.l., 1979-80).

FIELD CLIMATE CONTROL

In the late sixties, the thermoelectrically climatized gas
exchange chambers were developed by Koch and others
(1968). Hasler and Blaser (1981) measured the first daily
courses in young cembran pines on the east- and the
north-facing slopes of an avalanche gully at the timberline
at Stillberg. The two sites exemplify marked differences
in microclimate (mean values of soil surface temperatures
per growing season differ by more than 10 °C). These
differences in microclimate also £iffect the photosynthetic
rates of afforested cembran pines. Carbon dioxide gain,
summarized over a season, of the trees on the north-
facing slope was 34 percent less them for those on the
east-facing slope (Turner and Streule 1983).

Gas exchange measured in combination with meteoro-
logical parameters was registered in a branch of an older
cembran pine at Stillberg (1,980 m a.s.l.) throughout
a whole winter season in 1979-80 (fig. 1). Havranek
(1981) published an annual course of growth and a
carbon dioxide balance, taken near the Patscherkofel
laboratory. During summer the measurements were
taken with climatized chambers.

In the middle of the 1980's the so called "forest decline"
had high priority in forest research in Europe. The main
work in this field was done on other tree species than
cembran pine. Nevertheless, some investigations were
made on this species. Havranek (1987) took measure-
ments on cembran pine at the timberline to investigate
stress caused by air pollution in combination with climatic
factors. Genys and Heggestad (1978) did not find a high
susceptibility of cembran pines to ozone and sulphur di-
oxide, but MinarCic and KubiCek (1991) suggested that
dust could influence the stomata in needles of cembran
pines and thus affect photosynthesis. Lutz and others
(1988) emalyzed the photos)nithetic pigments of cembran
pines growing at different altitudes, which thus are ex-
posed to different ozone concentrations. Nebel and Matile
(1992) investigated the cause of uncommon yellowing of
the needles of cembran pines by analyzing photosynthe-
sis, nutrient relations, longevity, and senescence of these
declining needles.

The newest measm*ements in cembran pine from the
Stillberg experimental station are presented in these pro-
ceedings (Koike and others). These are probably the first
measvirements of photosynthesis in cembran pine at the
timberline with an eirtificially elevated atmospheric
concentration. A detailed publication giving these results
is in preparation (Koike and others 1993).

CONCLUSIONS

At the beginning of the research on cembran pine, about
150 years ago, scientists worked mainly at the forest
level. With the development of sophisticated measure-
ment instruments, branches, twigs, or even needles were
the focus of interest. Rarely did anyone try to extrapolate
these results from the small parts to a whole tree, a prob-
lem which has still not been satisfactorily resolved. Prob-
ably we should try to go back directly to the level of whole
trees or even of stands, especially when we want to under-
stand changes caused by external influences, such as air

pollution, rising carbon dioxide, changing UV radiation,
and many others.

ACKNOWLEDGMENT

I thank Mrs. M. J. Sieber for critical editing of the
English version of my manuscript.

REFERENCES

Baig, M. N.; Tranquillini, W. 1980. The effect of wind and
temperature on cuticular transpiration of Picea abies
and Pinus cembra and their significance in desiccation
damage at alpine treeline. Oecologia. 47: 252-256.

Baig, M. N.; Tranquillini, W.; Havranek, W. M. 1974.
Cuticulare Transpiration von Picea abies- und Pinus
cembra-Zvfeigen aus verschiedener Seehohe imd ihre
Bedeutung fiir die winterliche Austrocknung der
Baiune an der alpinen Waldgrenze. Centredblatt fiir
das gesamte Forstwesen. 91: 195-211.

Bamberg, S.; Schwarz, W.; TranquiUini, W. 1967. Influen-
ce of daylength on photosynthetic capacity of stone pine
(Pinus cembra L.). Ecology. 48: 264-269.

Caldwell, M. M. 1970. The effect of wind on stomatal
aperture, photosynthesis, and transpiration of Rhodo-
dendron ferrugineum L. and Pinus cembran L.
Centralblatt fur das gesamte Forstwesen. 87: 193-201.

Cartellieri, E. 1935. Jahresgang von osmotischem Wert,
Transpiration und Assimilation einiger Ericaceen der
alpinen Zwergstrauchheide und von Pinus cembran.
Jahrbiicher fiir wissenschaftliche Boteuiik. 82: 460-506.

Fankhauser, F. 1853. Uber das Vorkommen und
Gedeihen der Arve. Schweizerisches Forst Journal.
S83.

Fankhauser, F. 1903. Der Kiefemschiittepilz an der Arve.
Schweizerische Zeitschrifl fiir Forstwesen. 54: 321-323.

Genys, J. B.; Heggestad, H. E. 1978. Susceptibility of dif-
ferent species, clones and strains of pines to acute in-
jury caused by ozone and sulfur dioxide. Plant Disease
Reporter. 62(8): 687-691.

Gtinthardt, M. S. ; Wanner, H. 1982. Die Menge des
cuticularen Wachses auf Nadeln von Pinus cembran
(L.) und Picea abies (L.) Karsten in Abhangigkeit von
Nadelalter und Standort. Flora. 172: 125-137.

Hasler, R.; Blaser, P. 1981. Nettophotosjmtheseaktivitat
von Arve imd Larche auf verschiedenen Standorten der
subalpinen Stufe. Mitteilungen der forstlichen
Bundesversuchsanstalt, Wien. 140: 179-188.

Havranek, W. 1972. The influence of soil temperature on
photosynthesis and transpiration of young trees and on
dry-matter production, at the timber line. Angewandte
Botanik. 46: 101-116.

Havranek, W. M. 1981. Stammatmung, Dickenwachstum
und Photosynthese einer Zirbe {Pinus cembra) an der
Waldgrenze. Mitteilimgen der forstlichen Bundes-
versuchsanstalt, Wien. 142: 443-467.

Havranek, W. M. 1987. Physiologische Reaktionen auf
Klimastress bei Baumen an der Waldgrenze. Sympo-
smm Klima und Witterung in ZusEunmenhang mit den
neuartigen Waldschaden. Tagungsbeitrage, Miinchen:
115-138.

Havranek, W. M.; Benecke, U. 1978. The influence of soil
moistiire on water potential, transpiration and photo-
synthesis of conifer seedlings. Plant and Soil. 49(1):
91-103.

Holzer, K. 1958. Die winterlichen Veranderungen der
Assimilationszellen von Zirbe (Pinus cembra L. ) und
Fichte (Picea excelsa Link. ) an der alpinen Waldgrenze.
osterreichische botanische Zeitschrift. 105: 323-346.

Keller, C. 1890. Tierische Forstbeschadigiingen an der
Arve. Osterreichische Forstzeitung: 267-268.

Keller, C. 1901. Die Arven-Erkrankungen im
Oberengadin. Schweizerische Zeitschrift fur Forst-
wesen. 52: 293-299.

Keller, C. 1910. Die tierischen Feinde der Arve.
Schweizerische Centralanstalt fur forstliches. Versuchs-
wesen, Mitteilungen. 10(1): 1-50.

Keller, T. 1970. Uber die Assimilation einer jungen Arve
in Winterhalbjahr. Biindnerwald. 23(2): 49-54.

Keller, T.; Beda-Puta, H. 1973. Gas exchange of leafless
shoots in winter. Schweizerische Zeitschrift; fiir
Forstwesen. 124(6): 433-441.

Kemer, A. 1866. Wachstumsbedingungen der Zirbe.
osterr. Monatsschr. Forstwesen. 1-15.

Koch, W.; Klein, E.; Walz, H. 1968. Neuartige
Gaswechsel-Messanlage fiir Pflanzen in Laboratorium
und Freiland. Siemens Zeitschrift;. 42: 392-404.

Koike, T.; Hasler, R. ; Matyssek, R.; Item, H. 1993. Sea-
sonal changes in the photosynthetic capacity of Larix
decidua and Pinus cembran planted on contrasting
slopes at the timberline at Stillberg (Davos), eastern
Switzerland. [In preparation].

Kronfiiss, H. 1980. Das Bestandesklima einer
HochlagenaufForstung auf einem Sonnenhang. In:
Beitrage zur subalpinen Waldforschung. 2. Folge.
Mitteilvmgen der forstlichen Bundesversuchsanstalt,
Wien. 129: 81-103.

Kronfuss, H. 1983. Das Hohenwachstum einer Zirben-
aufforstung auf einem Siidhang in klimatologischer
Sicht. Allgemeine Forstzeitung. 94(12): 330-332.

Kronfuss, H. 1986. Die Abhangigkeit des Hohen-
zuwachses der Zirbe von der Seehohe. Jahresbericht
1985. Forstliche Bundesversuchsanstalt, Wien: 137-144.

Lutz, C; Schulte-Hostede, S.; Kirchner, M.; Reuther M.
1988. Photosjnithetische Pigmente aus Nadelbaumen
unterschiedlicher Hohenstufen des Otztals. Internatio-
nales Symposium ''Verteilung und Wirkimg von Photo-
oxidantien im Alpenraum. Sitzimgsberichte Miinchen
1988: 415-425.

MinarCic, P.; KubfCek, F. 1991. Localization of emissions
on the needle surface. 1. Methodology, morphological
characterization of the stomata {Picea abies, Pinus
cembran). Ekologia, CSFR. 10(4): 405-413.

Nebel, B.; Matile, Ph. 1992. Longevity and senescence
of needles in Pinus cembran. Trees. 6(3): 156-161.

Nevole, J. 1914. Die Verbreitimg der Zirbe in der osterr. -
ungar. Monarchic. Verlag W. Frick Wien. 89 p.

Pisek, A.; Winkler, E. 1958. Assimilationsvermogen und
Respiration der Fichte (Picea excelsa Lind. ) in verschie-
dener Hohenlage und der Zirbe (Pinus cembran L. ) an
der alpinen Waldgrenze. Planta. 51: 518-543.

Pisek, A.; Winkler, E. 1959. Licht- und Temperaturabhan-

gigkeit der COg-Assimilation von Fichte (Picea excelsa
Lind. ), Zirbe (Pinus cembran L. ) und Sonnenblume
(Helianthus annuus L. ). Planta. 53: 532-550.

Pisek, A.; Larcher, W.; Moser, W.; Pack, I. 1969.
Kardinale Temperaturbereiche der Photosynthese
und Grenztemperaturen des Lebens der Blatter ver-
schiedener Spermatophyten. III. Temperatur-
abhangigkeit und optimaler Temperaturbereich der
Netto-Photosynthese. Flora B. 158(3): 608-630.

Pisek, A.; Larcher, W.; Unterholzner, R. 1967. Kardinale
Temperaturbereiche der Photosynthese und Grenz-
temperaturen des Lebens der Blatter verschiedener
Spermatophyten. I. Temperaturminimum der Netto-
assimilation, Gefrier- und Frostschadenbereiche der
Blatter. Flora B. 157(1): 239-264.

Pisek, A.; Larcher, W.; Unterholzner, R.; Pack, I. 1968.
Kardinale Temperaturbereiche der Photosynthese und
Grenztemperaturen des Lebens der Blatter verschie-
dener Spermatophyten. II. Temperaturmaximima der
Netto-Photosynthese und Hitzeresistenz der Blatter.
Flora B. 158(2): 110-128.

Rikli, M. 1909. Die Arve in der Schweiz. Neue Denkschr.
Schweizerische naturforschende Gesellschaft. 44. 455 p.

Schonenberger, W.; Frey, W., eds. 1988. Untersuchungen
zur Okologie imd Technik der Hochlagenauffbrstimg.
Forschimgsergebnisse aus dem Lawinenanrissgebiet
Stillberg. Schweizerische Zeitschrift fiir Forstwesen.
139(9): 735-820.

Schwarz, W. 1970a. Der Einfluss der Photoperiode auf
das Austreiben, die Frostharte und die Hitzeresistenz
von Zirben und Alpenrosen. Flora. 159: 258-285.

Schwarz, W. R. 1970b. Frost hardiness, resistance to heat,
and COg uptake in seedlings and mature stages of
Pinus cembran L. Proc. 3. Forest Microclimate Sympo-
sium, Kananaskis Forest Experimental Station; 1969;
Seebe, AB: 137-145.

Schwarz, W. 1971. The photos)nithetic capacity of some
evergreens during the winter and the speed of their
recovery of activity after sharp frosts. Berichte der
deutschen botanischen Gesellschaft. 84(10): 585-594.

Simony, F. 1870. Die Zirbe. Jahrbuch des osterreichischen
Alpenvereins. 6: 349-359.

Tranquillini, W. 1955. Die Bedeutimg des Lichtes und der
Temperatur fiir die Kohlensaureassimilation von Pinus
cembran-tJungwuchs an einem hochalpinen Standort.
Planta. 46: 154-178.

Tranquillini, W. 1957. Standortsklima, Wasserbilanz und
COg- Gaswechsel junger Zirben (Pinus cembran) an der
alpinen Waldgrenze. Planta. 49: 612-661.

Tranquillini, W. 1958. Die Frostharte der Zirbe imter
besonderer Beriicksichtigung autochoner imd aus
Forstgarten stammender Jimgpflanzen.
Forstwissenschaftliches Centralblatt. 77: 89-105.

Tranquillini, W. 1959a. Die Stoffproduktion der Zirbe an
der Waldgrenze wahrend eines Jahres. Planta. 54:
107-129.

TranquiUini, W. 1959b. Die Stoffproduktion der Zirbe
(Pinus cembran L. ) an der Waldgrenze wahrend eines
Jahres. II. Zuwachs imd COg-Bilanz. Planta. 54(2):
130-151.

Tranqmllini, W. 1963a. Der Jahresgang der CO^-
Assimilation junger Zirben. Mitteilungen der forst-
lichen Bundesversuchsanstalt, Wien. 60: 501-534.

Tranquillini, W. 1963b. Die COg-Jahresbilanz und die
Stoifproduktion der Zirbe. Mitteilungen der forstlichen
Biindesversuchsanstalt, Wien. 60: 535-546.

Tranquillini, W. 1963c. Die Abhangigkeit der Kohlen-
saureassimilation junger Larchen, Fichten und Zirben
von der Luft- und Bodenfeuchte. Planta. 60: 70-94.

Tranquillini, W. 1963d. Uber die Frostresistenz der Zirbe.
Mitteilungen der forstlichen Bundesversuchsanstalt,
Wien. 60: 547-562.

Tranquillini, W. 1965. Uber den Zusammenhang zwischen
Entwicklungszustand und Diirreresistenz junger Zirben
{Pinus cembran L.) im Pflanzgarten. Mitteilungen der
forstlichen Bundesversuchsanstalt, Wien. 66: 241-271.

Tranquillini, W. 1979. Physiological ecology of the alpine
timberline. Berlin: Springer- Verlag. 137 p.

Tranquillini, W.; Machl-Ebner, I. 1971. Uber den Einfluss
von Warme auf das Photosynthesevermogen der Zirbe
(Pinus cembran L.) und der Alpenrose (Rhododendron
ferrugineum L.) im Winter. Report of the Kevo Subarc-
tic Research Station. 8: 158-166.

TranquilUni, W.; Schiitz, W. 1970. Uber die Rinden-
atmimg einiger Baume an der Waldgrenze. Central-
blatt fiir das gesamte Forstwesen. 87: 42-60.

Turner, H.; Streule, A. 1983. Wurzelwachstum iind
Sprossentwicklung junger Koniferen im Klimastress
der alpinen Waldgrenze, mit Berucksichtigung von
Mikroklima, Photosynthese und Stoffproduktion.
Irdning, Austria: Bundesanstalt Gimipenstein: 617-635.

Ulmer, W. 1937. Uber den Jahresgang der Frostharte
einiger immergriiner Arten der alpinen Stufe, sowie der
Zirbe und Fichte. Jahrbiicher fur wissenschaftliche
Botanik. 84: 553-592.

A PATTERN OF PINUS PUMILA SEED
PRODUCTION ECOLOGY IN THE
MOUNTAINS OF CENTRAL
KAMTCHATKA

Peter A, Khomentovsky

Abstract— This paper discusses the results of Pinus pumila cone
production studies in the mountains of central Kamtchatka. Pi-
nus pumila was found to have a very high seed production poten-
tial when not overtopped by neighboring trees. When shaded,
few or no cones were produced on P. pumila. Seed production
was not significantly dependent on altitude and site conditions.
However, elevations above 1,300-1,400 m were unfavorable for
cone survival, mainly because of low temperature and phenologi-
cal delay. Successful seed production requires snow cover to
protect ovulate buds on low-lying branches from freezing in the
winter. In the summer, good seed production is related to high
insolation, wind protection on cold sites, absence of shading, and
good drainage for the root systems. At midelevations, under good
seed production conditions in Kamtchatka, about 110 kg of seeds
per hectare will be produced as compared to only 1 kg at the up-
per elevational limits. Pinus pumila appears to have a high evo-
lutionary potential that reveals itself in prominent and wide
polymorphism of reproductive organs.

The main purpose of this paper is to give a brief descrip-
tion of some reproductive characteristics and ecological
features of typical subalpine dwarf pine (Pinus pumila
[Pall.] Regel) (Pinaceae, Strobus) (Pp) for many regions
of its area ecotopes of the central Kamtchatka Mountains
in northeastern Asia (fig. 1).

Some years ago, we tried to identify the response of
seed production (reflected in cones, seed, and nuclei masses
and sizes) to some of the most evident differences of Pp
environments, mainly to ecotopes, altitudinal position,
and their location in regions with maritime or continental
climate. We noticed (Khomentovsky and Khomentovskaya
1990) that there is no strong dependence of seed param-
eters and cone crop on the ecotope, elevation, and its posi-
tion in the relief, Pp was supposed to have some ecotopic
(macro- or microclimatic) optimum of seed production,
which could be indirectly characterized by the climate
within the belt of 600 to 900 m above sea level in the inte-
rior continental climate regions of Kamtchatka peninsula,
and widened in its lower limit almost to sea level near the
coast.

At the same time we tried to develop some of the first
descriptions of conophagous insects feeding on Pp and to

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Peter A. Khomentovsky is Head of the Biogeocenology Laboratory,
Kamtchatka Institute of Ecology and Nature Management, Prospect
Rybakov, 19-A, Petropavlovsk-Kamtchatsky 683024 Russia.

define their role in cone crop size and quality variation.
We found (Khomentovsky and Efremova 1991) that the
only two presently known insect species inhabiting Pp
cones are Cecidomyia pumila Mamaev and Efi-emova
(Diptera, Cecidomyidae) and Eupithecia abietaria Groeze
(Lepidoptera, Geometridae) (the taxonomy of the last spe-
cies has to be rechecked), and they affect crop quality very
little. Therefore, this damage could be treated as negli-
gible or at least not proven to be important.

These preliminary conclusions forced us to proceed with
a more detailed investigation of Pp seed production. Hav-
ing no possibilities for obtaining direct microclimatic in-
formation, we decided to define possible principal responses
of Pp seed production to elevation and ecotopical features
within a small creek basin with a high altitudinal gradient.

THE RESEARCH AREA

The research area is situated almost in the geographi-
cal center of the Kamtchatka (56 °N., 158 °E.), in the
Sredinnij Mountain Range system, the main system
in the peninsula (fig. 1). According to the botanical-
geographical regionalization scheme of L. F. Kimitsin
(1963), this territory is included in the "mountain above
alpine tundra shrubby region." According to oar scheme
(Khomentovsky and others 1989), it lies in the contact
area of "Central Kamtchatka plainland-submountain
province of conifer-stonebirch forests" and "Sredinno-
Western middle mountain stone birch-tundra forest prov-
ince." The area covers the upper part of the conifer forest
vegetation belt, subalpine, alpine, and above-alpine belts
in Kamtchatka (figs. 1 and 2).

Climate is subcontinental, with 300 to 400 mm of an-
nual precipitation, moderately cold and snowy winters,
and cool summers (monthly average air temperature in
Esso village is 3 °C in May, 9.5 °C in June, 13 °C in July,
12 °C in August, and 6 °C in September. The permanent
snow cover at elevations of 700 m and higher begins to
accumulate in the first half of October and disappears
mEiinly in May.

Orographically, the area can be called "typical middle
altitudinal-mountainous" (according to common under-
standing), which for Kamtchatka actually means middle
and high mountains (due to the compression of its altitu-
dinal vegetation belts). Average height of the mountains
is about 1,500 m. Geological and geomorphological struc-
ture of the area is determined by volcanogenic basalts and
andesits (Ql, Q2, Q4), and by quatenary glacial, fluvio-
glacial (Q3), alluvial, and proluvial (Q4) deposits. Glacial
relief is conspicuous, especially in the upper parts of river

Figure 1 — General description of the research area in Kamtchatka
peninsula. Top left: principal scheme of vegetation belts: 1 — subalpine,
2 — boreal forest, 3 — alpine, 4 — coastal and lowland. Top right: main
mountain ridges and hydrological net. Bottom left: research area, within
small basin. Dots are points where material was collected.

more or less permanent (on large time scale) weak or
moderate volcanic ashfalls impacts (mainly from the east-
ern part of peninsula).

Due to the geographical position, mountain relief, and
severe climate of the area, vegetation is not very diverse
and exhibits a mosaic-like spatial pattern. River valleys
and lower parts of their tributaries are occupied by Popu-
lus suaveolens, Chosenia arbutifolia, Larix cajanderi,

and creek basins. Soils are typical for the Kamtchatka
Mountain regions classification (I. A. Sokolov 1973), with
the additions by N. V. Kazakov (personal communication)
They include turf illuvio-humus, turf illuvio-volcanic ash,
tundra illuvio-humus, tundra illuvio-humus- volcanic ash,
and turf primitive. Soil genetics reveal themselves in the
existence of specific features including the long-term fi'eez-
ing period, anaerobic reactions in the lower horizons, and

Figure 2 — A fragment of the vegetation cover in the upper part of the altitudinal transect (see also fig. 1).

Betula kamtschatica, andB. ermanii. Two dwarf tree spe-
cies, Pinus pumila audi Alnus kamtschatica, are widely
spread along the whole vegetation profile. They occur by
groups or strips along the moraine tongues, river terraces,
fluvioglacial deposits, and watershed ridges and slopes,
and form the upper limit of woody vegetation (fig. 2).

Plant formations in the middle and high altitudes are
almost the same (except for Populus and Chosenia), but,
moving upward along the narrowing valleys and water-
sheds, along the altitudinal environmental gradient of
gradually decreasing plant life conditions, we can see
the distribution pattern of the formation changing from
"macromosaic" to "micromosaic." Tree distribution is
more strictly related to soil temperature conditions and
drainage gradients, and to sites sheltered from wind and
snow abrasion. This is true even for such ecologically
flexible dwarf species as Pinus pumila and Alnus
kamtschatica.

The upper elevational limits of upright-grovnng woody
vegetation {Larix kamtchatica, Betula ermanii) are lo-
cated between 900 to 1,000 m, depending on site condi-
tions. Pinus pumila and Alnus kamtschatica occur up to
1,300 to 1,400 m (rare cl\mips). Up to 1,200 m P. pumila
is able to produce seeds.

MATERIALS AND METHODS

Data were gathered mostly in September to October 1990
(with some additional data from other years) at the field
base "Bolgit," in the vicinity of Esso village (Bistrinsky
district of Kamtchatka region), in the "Tupikin clyutch"
creek basin, within an area of about 10 km^. This paper
presents the material from one slope exposed to the east-
northeast and from the area nearest to the watershed (up
to 50 m from the slope edge), within an altitudinal range
of 650 to 1,030 m. Permanent plots on an existing tran-
sect and some additional temporary points were chosen
for the cone crop measuring and sampling (fig. 1).

The follovmig information was collected on each sample
site (table 1): type of plant commimity, topographic

location and ground cover of Pp clumps, average seasonal
height of skeleton branches, general moisture conditions,
shading by neighboring upright trees or climips.

EXPLANATORY COMMENTS

First, strictly speaking, we cannot consider Pp plants
as common trees — they are not standing separately and
do not have a single trunk; in most cases they overlap
each other, both aboveground and in imderground parts.
Also, we cannot call them "clones" because we do not know
their origin and level of genetic relationships — which is
especially important in the case of seed dispersal by ani-
mals. The best term would be "clump" (we are grateful to
Dr. Diana Tomback for some explanations on this topic).

Second, the single trunk (in traditional understanding)
of Pp is considered a creeping tree not a bush, and its
length is only several centimeters or tens of centimeters.
Branches of first and second order (I prefer to call them
"trunk-branches") exhibit the same physiognomy, bear
similar quantity of shoots, and form multi trunk-like
crovms. In the case of syngenetic origin and continuous
dispersal, Pp plants form such dense cover by overlapping
basal parts in the litter and root systems that it is impos-
sible to identify single plants and thus their total num-
bers without destroying the stand. This forced us to use
the only acceptable method of measuring cone crops or
similar parameters — instead of measuring only one tree
or clump, measure by square unit such as hectares or
square meters. Certainly, the stand structure and charac-
teristics of groimd cover have to be taken into account,
especially for practical taxation.

Third, it is also impossible, in many cases, to determine
real age of adult Pp plants. They produce adventitious
roots, are growing apically, and are gradually decaying
in their basal parts during most of their lifetime, theoreti-
cally endlessly. We have to recognize that we really can
only know the age at the moment of measuring, similar
to being counted in young trees, and not the real one,
which is usually higher.

Table 1 — Short description of the ecotopes and Pinuspumila clumps being analyzed. The age of clumps in these analyses lies between 150
and 260 years'

Altitude,

Pinus pumila

Way of
main water

Woodstand

Pinus pumila

Shading

m asl

Position In the relief

community type

supply

composition

H avg, cm

SO, %

UTr

Ppc

1,030

Upper part of the flat
watershed, shaded from
S and E by the ridge

Pumilae pinetum
carioso-hypnoso-
ericosum, with
fragments of P.p.
cladinosum (CHE*C^

A+S

lOPp

100 (15-20)

80 unev

950

Plateau, opened to all
directions, watershed
with slight slope to
NNE

P.p. purum, with
fragments of
P.p. carioso-
cladinosum (P*CC^

lOPp

40 (10-15)

40 unev

900

Flat watershed above
the creek source,
shaded by the ridge
from S

P.p. hypnoso-

carioso-ericosum

(HCE)

A+S

lOPp+Lk
(upper
limit of
Lk)

300

40 unev

810

A ridge of the complex
watershed, slight
slope to the NE

P.p. carioso-

hypnoso-ericosum

(CHE)

lOPp+Lk

150 (35-40)

60 unev

800

Middle part of the E
exposed slope of wide
creek valley

P.p. hypnoso-

carioso-ericosum

(HE)

S+A

lOPp+Lk

200

680

Eastern border of the
watershed ridge with
flattened top

P.p. hypnoso-

carioso-ericosum

(HCE)

A+S

7Pp 3Lk

300

60 unev

650

Lower part of E
exposed slope in
narrow creek valley

P.p. ericoso-
sphagnosum

lOPp+Ak

300 (40-45)

100

^ Pinus pumila community type: in parentheses — abbreviation for fig. 8; Way of main water supply: A — atmospheric, S — slope; Woodstand composition: Pp — Pinus
pumila, Lk — Larix cajanderi, Ak — Alnus kamtschatica (dwarf alder); Pinus pumila: H avg — average height (cm), in parentheses — rough estimation of annual shoot
elongation (mm); SC% — surface covering by the clumps, "unev" — uneven; Shading: Utr— from upright trees of different species, Ppc — from Pinuspumila neighbor-
ing clumps; 0 — no shading, 1 — slight, 2 — moderate, 3 — heavy shading.

In each point for data and sample collecting v^^e used
3 to 5 small randomized plots of different size — from 1 to
15 m^. The following parameters were countered in place
(table 2): number of skeleton branches ("trunk-branches"),
number of germinating shoots on each skeleton branch;
number of current year crop cones (the second year of
cone development) and number of next year crop cones
(female buds of the first year) on each germinating shoot
and each skeleton branch; and among the total quantity of
each year's cones the number of cones damaged by insects
iCecidomyia pumila — damage in the first year of cone de-
velopment, Eupithecia abietaria — damage in the second
year of cone development), and by the nutcracker {Nuci-
fraga caryocatactes kamtschatkensis) .

At each site, 10 to 100 cones of the current year's crop
were collected and measured before they dried (cone
length and cone diameter). Later, after air drying during
some weeks or months, cones and seeds were measured
by 16 other parameters, six of which are discussed in this
paper: cone mass; 1,000 seeds mass; 1,000 nuclei mass
(both by measuring a number of samples with 50 seeds in
each); nimiber of seed scales in the cone; among the total

number of seed scales, the number of scales not contain-
ing seeds under them; and total seed quantity in the cone.

RESULTS AND BRIEF COMMENTS

• The structure of the cone-bearing portion of the
crown and cone crop estimates (tables 1 and 2, figs. 1-6):

As a whole, the number of skeleton branches and germi-
nating shoots per hectare varied very little within alti-
tudes of 650-800 m and definitely increased upward. This
could be explained by increasing isolation. Most likely it
was the reason for the especially increasing number of
germinating shoots (two to three times higher than skel-
eton branches) and, correspondingly, the nvimbers of
current-year and next-year cones. The last varied more
than the first because some of them vdll inevitably die in
the natural selection process during the maturation time
in the coming year.

The number of shoots with cones per skeleton branch
(fig. 4) changed opposite to that of the number of skeleton
branches per 1 hectare (we noticed the increased number

of shoots, including nongerminating). Even a slight plant
shading immediately resulted in a change in the number
of shoots (sites 900 m and 1,030 m).

The quantity of current and next-year crop cones per
one germinating shoot had specific dispersal features:
at each site, especially in the lower, shaded part of the
elevation profile, cones of any one year were overwhelm-
ing. This migration of intensive seed-producing centers
from year to year supports our previous conclusion

(Khomentovsky and Efremova 1991) about the existence
of local seed production mosaics, which provides a con-
tinuous supply of seeds, important not only for guaran-
teed reproduction and microevolutionary diversity for Pp,
but also for all zooconsumers such as nutcracker, other
birds, and mammals.

• Cones and seeds mass and size variation (tables 1, 3;
figs. 7 and 8):

Table 2 — Characteristics of germinating crown parts and cone crop outcome of Pinus pumila in central Kamtcliatka Mountains^

Parameter

Altitude (m above sea level)

650

680

800

810

900

950

1,030

No. of skeleton branches/ha

1,200

4,167

2,240

21,111

11,000

30,000

10,333

No. of germinating shoots/ha

1,733

7,917

3,200

47,778

57,000

86,667

33,000

No. of this year crop cones/ha

1,733

4,375

3,136

2,389

46,170

56,334

28,000

No. of next year crop cones/ha

260

160

51,122

19,950

143,000

24,550

No of aerminatina shoots/skel bran

1.44

1.90

1.43

2.26

5.19

2.89

3.00

No of thi^ vppir oroo conp^/<>kpl hri^n

1.44

2.00

1.40

0.11

4.20

1.88

2.56

No of next vear croo cones /skel bran

0.22

A A7

O AO

1 Md.

4.77

2.54

o
O

No. of this year crop cones/germ, shoot

1 .00

H AC
1 .00

0.98

0.05

0.81

A 7ft

No. of next year crop cones/germ, shoot

0.15

0.05

1.07

0.35

1.65

0.86

Percentage of skeleton branches with

various numbers of germinating shoots:

with one shoot

with two shoots

with thrfifi shoots

will 1 11 II Ol Iwwlw

with four shoots

with fiv/o chnr^tQ
Will i live? Ol IwUlO

with qIy QhnntQ

Willi OlA Ol lUUlO

A
0

with seven shoots

with eight shoots

with nine shoots

with ten shoots

with eleven shoots

with twelve shoots

with thirteen shoots

1 1

r ciccf iictgc ui gcniiiriaurig biiouis Dcaiirig

various number of this year cones (T)

and next year cones (N):

0N-1T

100

ON -21

Cumulative percent of shoots with

this year's crop cones only

100

1N-0T

2N-0T

3N-0T

Cumulative percent of shoots with

this year's crop cones only

1N-1T

1N -2T

2N - 1T

2N-2T

3N-1T

4N - 1T

Cumulative percent of shoots with

cones of both years

'Only the area covered with P'mus pumila clumps is taken into account (par. 1 -4), not the whole territory of the site; number of next-year cones can decrease
during the year due to natural selection processes.

Table 3 — Some characteristics of Pinus pumila cones and seeds in various elevations in central
Kamtchatka Mountains (within research area)'

Regression on altitude Correlation Average value

= 54.71

0.020 A

= -0.73

AVG (cl) =

+/-

8 (25-62)

= 33.68

0.010 A

= -0.79

AVG (cd) =

+/-

2(18-37)

= 7.81

0.030 A

= -0.41

AVG (cm) =

+/-

1 (4-10)

SSN

= 32.30

0.010 A

= 0.53

AVG(ssn) =

+/-

5 (26-52)

SWS

= 9.06

0.004 A

= 0.33

AVG (SWS) =

+/-

3 (5-18)

SQT

= 28.50

0.003 A

= 0.76

AVG (sqt) =

+/-

5 (26-68)

SMP

= 46.71

0.002 A

= -0.11

AVG(smp) =

+/-

3 (32-55)

SMTH

= 101.67

0.020 A

= -0.45

AVG (smth) =

+/-

8 (52-116)

NMTH

= 56.49

0.020 A

= -0.49

AVG (nmth) =

+/-

4 (27-68)

'CL — cone length, mm; CD — cone diameter, mm; CM — cone mass, g; A — altitude, m above sea level; SSN — seed
scale number; SWS— number of scales without seeds (within SSN); SQT — total number of seeds in the cone; SMP —
percentage of seeds mass in cone mass; SMTH — mass of 1 ,000 seeds, g; NMTH — mass of 1 ,000 nuclei, g. In paren-
theses at the right side — minimum and maximum values.

Mass variation (here and later we speak about current-
year crop cones only) is more evident in the upper part of
the elevation profile, in micromosaic sites with increasing
abiotic environmental pressure. In sufficiently insolated
(sunny) and wind-protected sites cone mass is increasing,
in shaded sites or windy plateaus — decreasing. At the up-
per elevation levels of Pp distribution (1,200-1,300 m) we
observed natural abrupt decreases of all seed-producing
parameters. It is remarkable that in general cones and
seeds masses are not too closely related to definite altitu-
dinal or site positions of the tree, even in cases of evi-
dently unfavorable envirormients (site 950 m).

Self regulation of seed production is obvious also: along
the elevation gradient, number of seed scales in the cone

(with and without seeds) and total number of seeds in the
cone slightly increases as well, and at the same time seed
and nuclei mass is unchanging or slightly decreasing. It
means that seeds become smaller but grow in quantity,
keeping the same reproduction potential and having more
protection from environmental impacts.

Cone size variation, most evident in their length, corre-
sponds vrith mass variation in relation to site position in
the landscape. Here, as well as in the previous case, the
main impact factor is not elevation but landscape struc-
ture of the ecotope (slopes in wide or narrow valley, pla-
teau, watershed ridge, etc.). Biocenotic structure is not
too important in this case, at least at the forest type level.
All ecotopes, excluding the 950-m site (opened plateau),

160

140

§ 120
O

£ 100

X 80

O 60

n
E

Z 40

i
i

— \

— i-/-

. \

/
/I

\ \
\\

\ 1

• y

>'
/

\l
1 •

1 .

1 /

1 ^

skeleton branches

germinating shoots

— current year cones
next year cones

650 680 800 810 900 950 1.030

Altitude (m as!)

Figure 3 — Principal structure of P'mus pumila crowns at
various altitudes and estimated cone crop per hectare.

germ, shoots/sk. br.

650 680 800 810 900 950 1,030

Altitude (m asl)

Figure 4 — Structure of Pinus pumila crown germinating
part (number of germinating shoots and estimated cone
crop per skeleton branch).

are occupied by the same group of forest types — Pumilae
pinetum carioso-hypnoso-ericosum, and its variations.
Generally speaking, cones size, as well as cones mass,
which has good correlation with them (r = 0.7-0.8), vary
within limits already known for the species.

• Damage caused by insects and birds (fig. 9):

This was investigated in some detail previously
(Khomentovsky and Efremova 1991) and is not likely

to effect any substantial change in average cone sizes: it
does not affect cone diameter and leads to cone length de-
creases up to 3-9 percent (Cecidomyia pumila). However,
in some cases this insect species damage results in under-
development of up to 20 percent of the seeds, but this im-
pact usually does not stop the development of the rest of
the seeds in the cone and is compensated at the popula-
tion level by the abundant production of seeds.

100

cones of both years
next year cones only
current year cones only

650 680 800 810 900 950 1,030

Altitude (m asl)

Figure 6 — Distribution of current-year and next-year cone crop on
germinating shoots.

The level of damage caused by Cecidomyia pumila,
clearly noticeable visually by the cone curving, is usually
rather high everywhere but it varies with different site,
environmental, and weather conditions, to which insects
are more sensitive. In our case it was changing from mod-
erate damage of 25-50 percent on the 650-m site (slope
in the narrow valley), the 800-m site (slope in the wide

valley), and the 900-m site (wind-protected watershed at
the upper limit of Larix) to a maximum of 100 percent on
the 810-m site (watershed ridge) and a minimal 8 percent
on the 1,030-m site (weakly wind protected and too cold
for insects ecotope).

Confirmation of the negligible character and amount of
insect damage is also seen in the level of seed extraction

o
m

<u
a>
0)
o
in

<a
c
o
O

650

800

900

1,030

CM
II

(/)

Altitude (m asl)

cone mass
SM std

CM std

50 nuclei mass

50 seeds mass
NMstd

Figure 7 — Variations in Pinus purpila cones, seeds, and nuclei mass.
SL — slopes, P — Plateau.

650 680 800 810 900 950 1,030

Altitude (m asl)

Figure 8 — Variations in Pinus pumila cone size. 1 — forest type (see table 1).

by the nutcracker, which collects only full, matured seeds
(fig. 9).

DISCUSSION AND CONCLUSIONS

Data analysis results, combined with some previous
material, allow us to make some conclusions about Pinus
pumila seed production in Kamtchatka, keeping in mind
the fragmentary nature of our study, even though it was
done in typical environmental conditions.

1. The main conclusion is that P. pumila seed produc-
tion has such a high potential in time and space that we
can speak about the environmental (abiotic and natural
biotic) impact on it in the frame of modification only but
not in the frame of regulation. Environmental stresses
usually cannot cause a crop crisis — trees have compensat-
ing mechanisms both at the organismic and population

level, and one of them is the microhabitat mosaics of an-
nual seed production.

2. Pinus pumila seed production is not (in Kamtchatka,
at least) significantly dependent on altitude and site con-
ditions. Landscape conditions (above facial level) are the
most important. However, this is true only when trees
are not overshadowed by neighboring upright trees; that
is, when Pp forms an independent vegetational belt in
more or less solid segments — in plain or mountain tundra,
upper forest-tundra ecotone, on the seashore dunes, in
wide river valleys, etc. Pp often grows as the lower canopy
of more or less dense Larix forests (more than 30 percent
of Larix cover), which have the same age and the same
syngenetic origin (very often in pyrogenetic ecotopes).
These shaded dwarf pine usually do not produce seeds
and develop vegetatively only.

3. The utmost upper limits of distribution (1,300-1,400 m
above sea level) are unfavorable for Pp seed production.

650

680

800 810 900

Altitude (m asl)

950

1,030

Ea+Nc

□

Ea+Cp+Nc

Figure 9 — Percentage of cones damaged by insects and birds. Cp — Cecidomyia
pumila; Ea — Eupithecia abietaria; Nc — Nucifraga caryocatactes.

mainly because of low temperature and phenological de-
lay. But even here, some cones can be found, and some-
times the crop can be abundant in wind-protected habi-
tats. Phenological delay often prevents the nutcracker
from collecting seeds here. These seeds, having been dis-
persed by birds from lower sites, and having more stable
genotypes, provide a better chance for successful refores-
tation in new areas. At the same time, dwarf pine regen-
eration at high altitudes is more or less provided by the
remaining abundant seeds. Besides this, numerous zoo-
consorts (mainly micromammals) will have a good food
supply most valuable in extreme conditions.

The same unfavorable conditions for seed production
can be found in the Pp belt at the seashore (Pacific
Kamtchatka coast), on the dunes nearest to the water.
But here, as well as in the mountains, trees utilize each
opporttmity for maximal realization of their reproduction
ability. Pp seed production on the second to third dune
of the coast belt is equal in cone size and mass parameters
to those at the 900- to 1,000-m altitude and in seed and
nuclei parameters to those at the 650- to 750-m altitude
in the continental climate portion of the peninsula.

Conditions moderately favorable for Pp seed production
can be found in plain and submoimtain (foothill) places,
for instance on gravel-sandy fluvioglacial or proluvial
deposits, where abundant cone crops sporadically occur.

The most productive seed zone lies in the vegetation
belt within the 600-900 m above sea level area. Here,
cone crops are mostly permanent and stable from year
to year.

4. Successful seed production of Pp needs the following
conditions: in winter — sufficient snow cover to protect buds
of low-ljdng trees from freezing; in summer — high insola-
tion, good wind protection on cold sites, absence of shad-
ing, and good drainage for the root systems. The main
adaptive feature of Pp is its ability to lay close to the
ground in the autumn. This process is initiated by low
temperature in the autumn before the snow cover. This
unique feature provides good chances of survival under
very extreme conditions. Strict dependence of Pp on snow
cover depth is proved by coincidence of its area border with
the snow depth isoline of 40 cm and more (Lukitcheva
1964; Tstcherbakova 1964).

5. Special research and preliminary work, including
mapping of various productivity sites, is needed for the
practical (silvicultural) estimation of current or future
cone crops of Pp. It is possible to show, as an example,
the scope of crop diversity. In the middle elevation site
(700-800 m), with 5 g of seed mass in the cone, 80 percent
of germinating shoots, and 80 percent of Pp covering the
ground, we can get about 110 kg of seeds per hectare. At
the upper limit of seed production, with 2 g of seed mass
in the cone, 30 percent of germinating shoots, 10 percent
of covering, we can get about 1 kg of seeds per hectare.
High variability of cone crop is illustrated in table 1 data:
taking into account only pure stands of Pp (without con-
sidering its real distribution on the surface) we can see
that the crop size on the 650-m site is about 6.5 kg/ha,
and about 212 kg/ha on the 950-m site.

6. Starting from the elevation that corresponds to the
upper limit of upright trees distribution (900 to 1,000 m)

in Kamtchatka, Pp produces an increasing number of
shoots. This can be interpreted as a organismic compen-
sation for the increasing severity of abiotic environmental
conditions. Photosynthetic ability increases, seed produc-
tion potential, as we saw, remains at the same high level,
and seed protection (thickness of seed walls) becomes
stronger. All this takes place mostly in the subalpine belt
and proves Pp belongs mainly to this type of vegetation.
Coincidence of these activated processes with the upper
limit of upright trees distribution allows us to hypothe-
size the existence of some kind of temperature threshold-
trigger. It could be the sum of effective temperatures,
above which the development of upright tree forms is pos-
sible, and below which only prostrate forms develop.

7. We agree with the opinion that Pp, as well as other
stone pines, has a high evolutionary potential, revealing
itself in prominent and wide polymorphism of reproduc-
tive organs and their functional flexibility (Pravdin and
Iroshnikov 1982). This is well illustrated by our work
and other sources. Ecotypical features are also evident,
accompanied by stability of the main seed-production
parameters. In fact almost all of this is similar to that
which is known from other Pp regions (Bobrinev and
Rylkov 1984; Efremova and IvHev 1972; Kapper 1954;
Khomentovsky and Khomentovskaya 1990; Krylov and
others 1983; Rush 1974).

8. Pp seed production potential exceeds the possibilities
of its realization in our research area, and most likely, in
the Kamtchatka peninsula as a whole. Ecotopical diver-
sity in Kamtchatka is evidently only a part of the whole
scope of acceptable conditions for the species (here we
can admit that low temperatures affect vegetative growth
three to four times more than they affect generative or-
gans variability). Therefore, we should look for possible
phylogenetic roots of this species in the regions with a
wider geoecological spectrum of sites.

9. Our observations (partly presented in this paper) al-
low us to speculate that Pp is a middle-altitudinal species,
with more preferable growth conditions in the moderately
continental climate (in relatively humid mountains) than
at the seashore. But, regarding the whole area, the defi-
nite and necessary influence of Pacific wet air masses re-
sults in considerable snow cover.

Combining this conclusion with some known hypothe-
ses of Pp evolution (Sotchava and Lukitcheva 1953;
Tikhomirov 1949; etc.) we think that this dwarf tree spe-
cies most likely appeared in Tertiary times in some cli-
matically subcontinental mountain regions of Angarida
(northeastern Asia). In Kamtchatka, pollen data show
Pp has been here for 1-1.5 million years (Malaeva 1967;
Tchelebaeva and others 1974). During Pleistocene inter-
glacial periods, and in the early middle Holocene, Pp
occupied periglacial zones and other places unacceptable
for more-thermophil tree species. The highest incidence
of polyembriony and underdevelopment of seeds in
Kamtchatka support the possible centers of origin in the
moderate continental climates when compared to Magadan
(northeastern continental part of Asia) and Buryatia
(southern part of eastern Siberia) (Iroshnikov 1972; Rush
1974). These features can be considered a reaction to less
favorable conditions of the secondary environments.

including Kamtchatka peninsula. According to A. I.
Iroshnikov (1972), Pp is one of the most well-adapted
species to the continental climate of northeastern Asia.

REFERENCES

Bobrinev, V. P.; Rylkov, V. F. 1984. Seed production of
dwarf stone pine. Inform, bull, of Tchitinsky Informa-
tion Center, #76-84. Tchita. 5 p. [In Russian].

Efremova, L. S.; Ivliev, L. A. 1972. On dwarf stone pine
seed producing at Kamtchatka. In: Utilization and re-
producing of forest resources in the Far East. Abstr. of
the all-imion conference Khabarovsk. Part 2: 158-159.
[In Russian].

Iroshnikov, A. I. 1972. Seed producing and seed quality
of conifers in northern and mountainous regions of
Siberia. In: Seed producing of Siberian conifers (collec-
tion of papers). Novosibirsk: 98-117. [In Russian].

Kapper, O. A. 1954. Conifers (forestry characteristics).
IVloscow-Leningrad. 304 p. [In Russian].

Khomentovsky, P. A.; Efremova, L. S. 1991. Seed prod-
uction and cone-feeding insects oiPinus pumila at
Kamtchatka peninsula: aspects of coexistence. In:
Baranchikov, Y. N.; Mattson, W. J.; Hain, F. P.; Payne,
T. L., eds. Forest insects guilds: patterns of interaction
with host trees. U.S. Department of Agriculture, Forest
Service, Gen. Tech. Rep. NE-153: 316-320.

Khomentovsky, P. A.; Kazakov, N. V.; Tchernyagina, O. A.
1989. Tundra-forest zone at Kamtchatka: problems of
conservation and management. In: Problems of nature
management in taiga zone (collection of papers).
Irkutsk: 30-46. [In Russian].

Khomentovsky, P. A.; Khomentovskaya, I. G. 1990. Geo-
graphical variability of Pinus pumila seed producing at
Kamchatka. In: Questions of Kamchatka geography.
Petropavlovsk-Kamchatsky. No. 10: 47-55. [In Russian].

Krylov, G. V.; Talantsev, N. K; Kozakova, N. F. 1983.
Siberian pine (Pinus sibirica Du Tour). Moscow. 216 p.
[In Russian].

Kunitsin, L. F. 1963. A pattern of Kamtchatka nature
regionalization. In: Nature conditions and regionaliza-
tion of Kamtchatka district. Moscow: 7-26. [In Russian].

Lukitcheva, A. N. 1964. Distribution areas of Asia plants
(a map). In: World geographical atlas. Moscow. 112 p.

Malaeva, E. M. 1967. Kamtchatka vegetation develop-
ment in Pliocene-Pleistocene. In: Development of Sibe-
rian and far-eastern vegetation in Quaternary (collec-
tion of papers). Moscow: 78-170. [In Russian].

Pravdin, L. F.; Iroshnikov, A. I. 1982. Genetics of Pinus
sibirica Du Tour, P. koraiensis Sieb. et Zucc. and P.
pumila Regel. Annales Forestales. 9(3): 79-123.

Rush, V. A. 1974. Biochemical characteristics of five-
needle pines seeds. In: Biology of conifer seed repro-
duction in western Siberia (collection of papers).
Novosibirsk: 180-184. [In Russian].

Sokolov, I. A. 1973. Volcanism and soils formation
(with Kamtchatka as an example). Moscow. 224 p.
[In Russian].

Sotchava, V. B.; Lukitcheva, A. N. 1953. On geography
of dwarf stone pine. Reports of the USSR Academy of
Sciences. 90(6): 1163-1166. [In Russian].

Tchelebaeva, A. I.; Shantser, A. E.; Egorova, I. A.; Lupikina, E.
G. 1974. Cenozoic deposits of the Kuril-Kamtchatka
area. In: Kamtchatka, Kuril and Commander islands
(series of History of the relief development.) Moscow:
31-57. [In Russian].

Tikhomirov, B. A. 1949. Dwarf stone pine {Pinus pumila
(Pall.) Rgl), its biology and utilization. MOIP series,
botan. division, Moscow. 14(6): 106 p. [In Russian].

Tstcherbakova, E. J. 1964. An average of maximal decade
snow depths (a map). In: World geographical atlas.
Moscow. 220 p.

NEEDLE LONGEVITY AND
PHOTOSYNTHETIC PERFORMANCE IN
CEMBRAN PINE AND NORWAY SPRUCE
GROWING ON THE NORTH- AND EAST-
FACING SLOPES AT THE TIMBERLINE
OF STILLBERG IN THE SWISS ALPS

Takayoshi Koike
Rudolf Hasler
Hans Item

Abstract — The number of shoots in each age class and the pho-
tosynthesis of the shoots were measured on cembran pine {Pinus
cembra) and Norway spruce (Picea abies) growing on north- and
east-facing slopes at 2,185 m above sea level. Needle longevity
on the north-facing slope exceeded needle longevity on the east-
facing slope by 1 to 3 years in both species. A negative correla-
tion between maximum photosynthesis and needle lifespan was
foimd. However, there was no physiological difference in the
photosynthesis of trees on the respective slopes.

In the Swiss Alps, cembran pine (Pinus cembra) grov^rs
around timberline level and Nonvay spruce (Picea abies)
from valleys to mountainsides (Kuoch and Amiet 1970).
At the subalpine timberline, the growth and biomass in-
crease of trees are strongly influenced by the microsite
performance and microclimate (Livingston and Black
1987; Schonenberger and Frey 1988; Tvu-ner and others
1982). Using the monitoring experiment on photosyn-
thetic production as a basis, the biomass increase of
mountain pine (Pinus mugo) growdng at the timberline
was found to be regulated by the soil temperature and the
net radiation (Hasler 1982). However, physiological char-
acteristics of trees were not revealed. Is the turnover rate
of needles related to the tree size? What about the physi-
ological adaption of needles to the microclimate?

In this study, these questions, needle lifespan and the
age-related photosynthesis of cembran pine planted on
north- and east-facing slopes, were measured. We present
the physiological parameters of needle photosynthesis.
Anatomical traits of needles were investigated with refer-
ence to the estimation of the intercellular CO^ concentra-
tions (C.) (Terashima 1992). We discuss the trade-off rela-
tion between the needle lifespan and photosynthetic rates
influenced by the microclimate.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Takayoshi Koike, Rudolf Hasler, and Hans Item are members of the
"Ecophysiology" group, Swiss Federal Institute for Forest, Snow and Land-
scape Research, CH-8903 Birmensdorf, Switzerland. Present address of
T. Koike: Forest Products Research Institute, Sapporo 062, Japan.

MATERIALS AND METHODS

At the study site (Stillberg in the Swiss Alps), on the
north-facing slope, the yearly net radiation surplus was
308 kWh/m^ lower, and the root-zone temperature during
grov^rth period was 2-4 °C lower than on the east-facing
slope at 2,185 m above sea level (a.s.l.) (Turner and others
1982). The mean soil temperature at 10 cm on the north-
and east-facing slopes from June to September was 3.5
and 4.5 °C, respectively (Koike and others, submitted).

Plant Materials

Plants studied were cembran pine, (provenance 2,050 m
a.s.l.) and Norway spruce (provensmce 1,960 m a.s.l.),
raised from seeds in the nursery at Birmensdorf ZH (550 m
a.s.l.). The cembran pines were transplanted in 1975,
and the Norway spruces in 1962, to the experimental site
(2,185 m a.s.l.). Mean tree height as of 1988 on the north-
facing slope and east-facing slope was 40 cm and 80 cm in
cembran pine and 36 cm and 45 cm in Norway spruce, re-
spectively. For measurements of photos3mthesis, attached
shoots vdthin the even-aged needles were used. Younger
shoots were removed when the photosynthesis of older
shoots was measured. The cut shoot ends were sealed
with vaseline.

Measurements

The gas exchange rate was measured with a thermo-
electrically controlled chamber (Mini-cuvette Walz, FRG),
which was coated with aluminium foil in order to fully
use diffuse light (Koike and others, submitted). Photosyn-
thesis was determined v^dth two infrared gas analyzers
(Binos, Leybold-Heraeus, FRG). Dew point mirrors (Walz,
FRG) were used to measure the absolute air hmnidity.
Needle temperatures were monitored by a 0.1 mm chromel-
constantan thermocouple. The PPFD above the needles
was measured with a GaAsp photodiode (Hamamatsu
Gil 18, J) after calibration vsrith a LiCor quanttmi sensor
(USA). A COj dispensing apparatus (Walz, FRG) pro-
duced different COg concentrations in the chamber.

Figure 1 — Cross section of a needle of a cembran
pine (right) and a NoPA'ay spruce (left).

Anatomy of needles was observed with a microscope
after needles were sectioned with a microtome (Leitz).
Total needle surface area of the used twig was estimated
by the modified glass bead method (Davies and Benecke
1980). The number of specimens for the census of needle
longevity was more than 700 shoots of 10 trees. Dry
weight of the needles was determined after 48 hours dry-
ing at 80 °C. Nitrogen concentration in the needles was
analyzed with a CHN analyzer (Rapid, USA).

RESULTS

Needle Characteristics

The anatomy of needles of cembran pine and Norway
spruce was similar to the "homobaric" leaves (fig. 1).
There was no extension of vascular bundle sheaths into
the mesophyll. Needle lifespans of cembran pine and
Norway spruce were 9 and 12 years for the north-facing
slope and 8 and 9 years for the east-facing slope, respec-
tively (fig. 2). The proportion of older needles of both spe-
cies was larger on the north-facing slope than on the east-
facing slope. In Norway spruce, the frequency of younger
needles on the north-facing slope was smaller than that
on the east-facing slope. A negative correlation between
the Ught-saturated photosynthesis and needle longevity
was found in both species (fig. 3).

Physiological Parameters

There was a positive correlation between the stomatal
conductance and the light-saturated photosjoithesis in
both species (fig. 4). No difference between the north- and
east-facing slopes was foimd. The quantum yield and car-
boxylation efficiency (CE) of both species on the north-
facing slope tended to be lower than of those on the east-
facing slope (table 1).

The maximimi photo sjTithe sis at light and CO^ satura-
tion (Pj^^) of both species on the north-facing slope was
lower than that on the east-facing slope (P < 0.05). The
CE of cembran pine was sHghtly higher than that of
Norway spruce. The nitrogen concentration in needles
of cembran pine was higher than in needles of Norway
spruce. However, there was no statistical difference in
both species between the slopes.

DISCUSSION

The estimation of is based on the uniform response
of the stomata of a leaf (Terashima 1992). The anatomy
of needles of cembran pine and Norway spruce was simi-
lar to the homobaric leaf. Non-uniform stomatal aperture
in needles of loblolly pine was observed only when needles

Figure 2 — Frequency of needle age in percent of Figure 3 — Photosynthesis in relation to needle age

total needles of cembran pines (right) and Norway in cembran pines (left) and Norway spruces (right)

spruces (left) growing on east-facing and north- growing on east-facing and north-facing slopes at

facing slopes at timberline. timberline.

w
o

o
sz

-54

• North-facing slope

uusi-Taciny siop6

0 40 80 0 40 80

Stomotal conductance (mmolrfi^s^)

Figure 4 — Photosynthesis in relation to stomatal
conductance in cembran pines (left) and Norway
spruces (right) growing on east-facing and north-
facing slopes at timberline.

were treated with abscisic acid, not with the low tempera-
ture in roots (Day and others 1991). According to these

facts, we can estimate the CE calculated with C. values.

The P^^ from needles of the north-facing slope was
smaller than of those of the east-facing slope, while needle
lifespan of the north-facing slope increased. Many physi-
ological parameters in needles from the north-facing slope
were almost the same as those on the east-facing slope.
However, slightly lower P^^ in both species from the
north-facing slope may be attributed to the lower soil
temperature (Day and others 1991; Hasler 1982, Turner
and others 1982). The lower P of needles from the

max

north-facing slope may be compensated for by the longer
lifespan of needles. Tree growth on the north-facing slope
may be retarded by the shortage of radiation and low soil
temperatures.

ACKNOWLEDGMENTS

We thank Dr. H. Turner for providing the climate
data of the Stillberg site. Prof. W. K. Smith for collecting
references, and Mrs. M. J. Sieber for editing the English
version of this manuscript.

REFERENCES

Davies, C. E.; Benecke, U. 1980. Fluidized bed coating
of conifer needles with glass beads for determination
of leaf surface area. Forest Science. 26(1): 29-32.

Day, T. A.; Heckathorn, S. A.; DeLucia, E. H. 1991. Limi-
tations of photosynthesis in Pinus taeda L. (loblolly
pine) at low soil temperatures. Plant Physiology. 96:
1246-1254.

Hasler, R. 1982. Net photosynthesis and transpiration
of Pinus montana on east and north facing slopes at
alpine timberline. Oecologia. 54: 14-22.

Koike, T.; Hasler, R.; Matyssek, R.; Item, H. Seasonal
changes in the photosynthetic capacity of Larix decidua
and Pinus cembra planted on contrasting slopes at the
timberline at Stillberg (Davos), eastern Switzerland.
Trees, submitted.

Kuoch, R.; Amiet, R. 1970. Die Verjiingung im Bereich der
oberen Waldgrenze der Alpen. Eidgenossische Anstalt
fiir das forstliche Versuchswesen, Mitteilungen. 46:
159-328.

Livingston, N. J.; Black, T. A. 1987. Stomatal characteris-
tics and transpiration of three species of conifer seed-
lings planted on a high elevation south-facing clear-cut.
Canadian Journal of Forest Research. 17: 1273-1282.

Schonenberger, W.; Frey, W. 1988. Untersuchungen ziar
Okologie imd Technik der Hochlagenaufiforstimg.
Forschungsergebnisse aus dem Lawinenanrissgebiet
Stillberg. Schweizerische Zeitschrift fur Forstwesen.
139: 735-820.

Terashima, L 1992. Anatomy of non-uniform leaf photo-
synthesis. Photos)aithesis Research. 31(3): 195-212.

Turner, H.; Hasler, R.; Schonenberger, W. 1982. Contrast-
ing microenvironments and their effects on carbon up-
take and allocation by yoimg conifers near alpine tree-
line in Switzerland. In: Waring, R. H.: Carbon uptake
and allocation in subalpine ecosystems as a key to man-
agement: Proceedings of the lUFRO Workshop,
Pl.07-00, Corvalhs, OR: Oregon State University:
22-30.

Table 1 — Physiological parameters of cembran pine and Norway spruce growing on an east- and a north-facing slope of an
avalanche gully at Stillberg (2,185 m a.s.l.)

Parameter

Cembran pine

E slope

N slope

Norway spruce

E slope

N slope

Quantum yield 0.014 ±0.004
(mol COj'mol/quanta)

Carboxylation efficiency 0.024 ± 0.009
(^mol•m-2•^ba^^'

(mol»m-2»s-^)

8.30 ± 1.51

Nitrogen concentration
(mg»g-^) 12.26 ± 0.90

0.009 ± 0.002

0.022 ± 0.009

5.38 ± 0.77

13.27 ± 0.91

(1)

0.012 ± 0.006 0.006 ± 0.002

0.017 ± 0.002 0.018 ± 0.008

6.87 ± 1.61

10.43 ± 1.87

4.93 ±0.12

9.26 ± 1.22

(1)

'P<0.05.

THE BROAD-LEAVED KOREAN PINE
FOREST IN CHINA

Luo Ju Chun

Abstract — Korean pine (Pinus koraiensis) is a tree species of
high economic value. Its natural distribution area, the character-
istics of distribution in the broad-leaved Korean pine forest re-
gion of China, the biological and ecological characteristics of Ko-
rean pine, the forest types of Korean pine, and the succession
pattern of the Korean pine forest communities have been studied.
The results are significant to conservation and sustaining use of
Korean pine.

Korean pine (Pinus koraiensis) is well known for its
high-quality timber that can be used for many different
purposes. Undoubtedly, Korean pine is one of the most
important tree species in Chinese forests.

Except for a few pure Korean pine forests, the species
often occurs in the "conifer-broad-leaved tree mixed-forest
communities" vnth many species of deciduous broad-
leaved trees and other conifers, which is the most repre-
sentative vegetation formation in the northeastern humid
zone in China.

NATURAL DISTRIBUTION AREA

Korean pine — a relict species from the Tertiary —
mainly occurs in northeast Asia and in a narrow zone
along the west coast of the Japan Sea. This zone includes
the southern part of eastern Russia, the eastern part of
northeast China, North Korea, and the center of Honshu
in Japan. The main natural range of Korean pine, how-
ever, is in China. Korean pine is a representative species
in Dahurian flora, and also is the main forest tree species
in the eastern momitains of northeast China.

In China, the broad-leaved Korean pine forest is found
from 40° 15' N. to 50°20' N. and from 126° E. to 135° E.,
east of Song-nen plain and north of Song-liao plain. In
the south it reaches Dan dong, in the north Hei-Ho
coimty. The whole sirea takes a crescent shape covering
Lesser Xingan Ling, Wan Da Mountain, Zhang Guang Cai
Ling, Lao Ye Ling, and Chang Bai Mountain. Most of
these mountains are trending from northeast to south-
west. They are characterized by gentle topography and
low elevation, not higher than 1,300 m. There are many
different tree species, including many quite valuable
broad-leaved trees. Their exploitation began more than
100 years ago. Their timber output still makes up one-
fourth of total timber production in China.

The forest region, which is influenced by the Japan Sea,
is characterized by temperate-zone monsoon climate. The

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Luo Ju Chun is Professor of Ecology and Forest Management, Forestry
Resource College, Forestry University, Beijing, China.

warm and moist climate is favorable to growth of Korean
pine. Annual average temperature is 0-6 °C; the growing
season is about 4-5 months. Temperatvu"e svun (>10 °C)
is about 2,500-3,200 °C. At higher latitude, temperature
is low, vyinter is long, and simimer short. Winter covers
5 months. January is the coldest month; the absolute
minimum temperature is -40 to -30 °C. In July, which is
the warmest month, mean temperature reaches 20-24 °C.
Winter is characterized by thick snow cover and mean
temperatures ranging from -28 to -14 °C. Annual pre-
cipitation amoimts to 600-1,100 mm £ind decreases from
south to north; maximum is in summer (June to August).
The annual relative hiunidity is about 60-70 percent.
Hiunidity and temperature are favorable to growth of
Korean pine and other trees.

The soil in Korean pine forests is dark brown forest soil
vfith a thick hvunus layer, and it is very fertile.

CHARACTERISTICS OF
VEGETATION DISTRIBUTION

Horizontal Distribution

Because of the wide range of Korean pine forest, the cli-
mate is quite different from north to south in the distribu-
tion area. Consequently, some differences in flora and
plant commvuiities become apparent. The distribution
area can be divided into two subzones: northern and
southern temperate. The demarcation is from Don-g Nin
in the eastern part of Jing Bo Lake to Ji Lin district until
its west boiindary; it seems to nm along the parallel of
44 °N. The differences in natiiral conditions and vegeta-
tion can be described as follows:

1. In the northern subzone, climate is cold, the growing
season is only about 100-120 days, and annual rainfall
£imounts to 500-700 mm; the southern subzone is charac-
terized by a warm and hvunid climate and a growing sea-
son of 130-150 days. Annual precipitation is about 700-
1,100 mm.

2. In the northern subzone, more conifers typical of the
cold-temperate zone (for example, Picea koraiensis, P.
jezoensis, and Abies nephrolepis) occur in Korean pine
forests. Pinus pumila can be found in the subalpine. In
addition, Quercus mongolica is more common on expo-
siu-es to Bun. Larix gmelinii forests, mixed with Betula
platyphylla, occur in the lowlands. Vaccinium vitis-
idaea, Ledum palustre, Vaccinium uliginosum, Betula
fructicosa — representatives of Dahurican flora — are typi-
cal of larch forest undergrowth. In the southern subzone,
there are more representatives of Chang Bai flora, such as
Picea jezoensis var. komarovii, Abies holophylla, and
Larix olgensis.

3. In the southern subzone there are more plant species
than in the north. The forest often consists of 30-40 woody
species forming three layers. In the southern subzone
there are twice as many broad-leaved tree species in
Korean pine forests than in the northern subzone; the
genus Acer, for instance, is represented by only five spe-
cies in the north, compared to 12 species in the south.
The percentage of valuable broad-leaved trees (such as
Franus mandshurica, Juglans mandshurica, Phello-
dendron amurense, Tilia amurensis, Ulmus propingua)
in Korean pine forests is higher in the southern than in
the northern subzone. These broad-leaved trees usually
grow together with Betula costata and Acer mono and rep-
resent more than 50 percent of the species of southern
Korean pine forests. Moreover, Carpinus cordata and
Fraxinus rhynchophylla are very commom in the south,
but occur rarely in the north. In addition, there are more
than 10 vine plant species in the southern subzone, but
only five or six in the north.

1400 -r

0-^

Figure 1 — Vertical distribution pattern of veg-
etation in Lesser Xin An Ling (southern expo-
sures): 1a, River head-valley-conifer forest with
larch, spruce, fir; 1b, Bank depression, broad-
leaved forest with Ulmus propinqua, Fraxinus
mandshurica, Pliellodendron amurense; 2, Oak
forest {Quercus mongolica) mixed with Betula
dahurica, Tilia mandshurica; 3, Broad-leaved
Korean pine forest mixed with spruce, fir,
Betula costata, Tilia amurensis, Ulmus
laciniata; 4, Spruce-fir forest belt; 5, Subalpine
Betula ermanii-Pinus pumila elfin forest belt.

Vertical Distribution

The vertical distribution pattern of vegetation in Lesser
Xin An Ling (fig. 1) is a typical example of the northern
subzone:

1. River valley forest: It is composed of Picea jezoensis,
Abies nephrolepis, Larix gmelinii, Juglans mandshurica,
Ulmus propinqua, Populus ussuriensis, and other species.

2. Oak {Quercus mongolica) broad-leaved forest belt
(150-300 m): It is secondary forest that established after
destruction of virgin conifer-broad-leaved mixed forest,
and it occurs usually on shallow soils on s\in-exposed
slopes.

3. Broad-leaved Korean pine forest belt (250-650 m):
Korean pine is the dominant species; spruces and firs oc-
casionally occur; the broad-leaved trees make up less than
30 percent of the species. The volume of standing timber
is the highest (300-600 m^/ha) in the northern forest re-
gion. Korean pine grows on exposures to south and south-
west, on mountain ridges, and on hill tops.

4. Spruce-fir forest belt (650-1,000 m): Picea korai-
ensis, Picea jezoensis, and Abies nephrolepis are the main
tree species.

5. Subalpine moss-elfin forest belt (1,000-1,080 m):
The climate is cold and humid; wind is strong. The elfin
forest is composed of Betula ermanii. This species exhib-
its dwarf growth forms not taller than 4-6 m. Pinus
pumila grows in the forest understory. The forest floor
is covered mainly by moss.

The vertical distribution pattern of vegetation in the
southern subzone can be illustrated by the situation of
Chang Bai Mountain (fig. 2):

1. Broad-leaved forest belt (250-500 m a.s.l.): It is
formed by Quercus mongolica, Betula platyphylla, Betula
dahurica, and Populus davidiana. It is a typical secon-
dary forest following destruction of broad-leaved Korean
pine forest. Locally, basswood, maple, elm, walnut, and
corktree form mixed forests. At lower elevations, these
broad-leaved forests are scattered.

2. Broad-leaved Korean pine forest belt (500-1,200 m):
Korean pine is the main species, and is mixed with vari-
ous broad-leaved trees: Betula costata, Carpinus costata,
Ulmus laciniata, Acer mono, Tilia amurensis, Quercus
mongolica, especially the valuable broad-leaved tree spe-
cies Ulmus propinqua, Phellodendron amurense, Juglans
mandshurica, Fraxinus mandshurica, and, at valley bot-
toms, some spruce and fir can be found.

3. Spruce-fir forest belt (1,200-1,800 m): It is formed
by Picea jezoensis, Picea koraiensis, and Abies nephrolepis,
with additional Betula ermanii and Larix olgensis in the
upper part of the belt, or with Korean pine, Betula
costata, Ulmus laciniata, and Tilia amurensis in its lower
part.

4. Subalpine Betula ermanii mossy elfin forest belt and
subalpine meadow belt (1,800-2,100 m): These belts are
characterized by cold and windy climate, steep topogra-
phy, and shallow soils. Betula ermanii forests alternate
with vast meadow areas.

5. Alpine tundra belt (>2,100 m): Vegetation consists
of different dwarf shrub species, mosses, and lichens.

3000 -r

2500

2000 -

5 1600 -
0)

1000 -

500 -

0-^

Figure 2 — Vertical zonal spectrum of vegetation
in Chang Bai Mountain: 1 , Secondary deciduous
broad-leaved forest; 2, Conifer-broad-leaved
mixed forest (broad-leaved Korean pine forest);
3, Coniferous forest belt: 3a, Spruce-fir-Korean
pine forest subbelt; 3b. Spruce-fir coniferous for-
est subbelt; 4, Subalpine Betula ermanii-e\i\n for-
est belt, and subalpine meadow belt; 5, Alpine
tundra belt.

As far as we know, the higher the latitude the lower is
the upper limit of Korean pine. In the southern subzone
Korean pine occurs up to 1,200 m a.s.l.; in the north its
uppermost altitudinal limit is located at about 650 m.

BIOLOGICAL AND ECOLOGICAL
CHARACTERISTICS

Korean pine is a long-lived species. Its lifespan is 300-
400 years, sometimes 500 years. Tree height usually
reaches 30 m. The diameter can reach 1 m. Korean pine
is one of the valuable sawtimber trees in the world. It
grows slowly in the first 5-8 years. After the age of 10 its
growth rate increases rapidly. Its diameter growth begins
to accelerate at the age of 12. The height growth becomes
more rapid after 16-18 years. Accelerated growth contin-
ues vmtil the age of 30. Korean pine is a tree species
growing fastest at middle age. In general, growth rate of
Korean pine increases earlier in plantations than in natu-
ral forests. In clear cuttings, the radial growth of 8-year-
old trees is 2.6 times higher than in nattu-al forests, and
height growth is twice as much as in trees at the same
age that became naturally established.

Secondary growth of Korean pine is common in autiunn,
but it appears mainly in trees less than 10 years old.
After 80 years growth, Korean pine reaches a height of
18-20 m. At this age the tree top often begins to fork and
bear cones. Obviously, the forking is closely related to
cone production.

Korean pine has shallow roots. Its taproot is imderde-
veloped and degenerates easily. Thus, seed trees left on
clear cuttings are sensitive to windbreak.

Most Korean pine trees begin to produce cones at the
age of 80-100 years. Nevertheless, cones can also be
found on 30-year-old trees. In plantations, however,
even 20-year-old trees are able to bear cones.

There are two ecot5T3es of Korean pine: one is the
leptodermis form displaying thin bark with small and
shallow scale or long strip lobes. It forks less, grows
faster in height, and produces better timber. The other
is the pachidermis form characterized by deep and large,
long, square-shape lobes on the tree trunk. It forks more,
grows slower in height, and the timber quality is not as
good as in the leptodermis type. The leptodermis type
should be used for afforestation.

Korean pine belongs to thermophile species; its vegeta-
tive growth requires 6-7 °C of the lowest average tempera-
tiu-e and the most suitable temperature is 14-16 °C in the
yoimg stage. The relative moistvu^e it demands is 70 per-
cent and, during the period of most rapid growing, rela-
tive hvunidity of about 70 percent is required. Korean
pine is intolerant to high temperature during vegetative
growth. If average temperature exceeds 15-16 °C, height
growth will decrease or even stop. On the other hand, it
will not be affected by winter temperatiires as low as
-50 °C.

At seedling stage a little shading is favorable to Korean
pine, althovigh it can tolerate full siinlight. Its light re-
quirement increases with age. While 3 years old, seed-
lings require 60-70 percent of incoming solar radiation.
After the fifth year, young trees become totally shade
intolerant.

Korean pine grows best on thick, moist, fertile, and well-
drained soils with pH 5-6, but it can also exist on shallow
dry soil, poor in nutrients. It grows badly on soils rich in
clay. However, Korean pine requires appropriate soil mois-
ture. In bogs and on sites with stagnant water, growth is
hampered, and Korean pine is rarely seen there.

PRINCIPAL FOREST TYPES

Mu Dan Jiang forest region, the distributive center of
Korean pine, may serve as an example to classify Korean
pine forest into the following forest types:

Steep-Slope Carex calltnichos-Korean Pine
Forest — This type occurs on simny exposures or narrow
ridges at an altitude of 800-1,000 m. These usually are
poor sites. The conmiunity can be divided into two layers.
The forest canopy is formed by Korean pine (90 percent)
and Quercus mongolica or Tilia amurensis (10 percent),
which have a higher crown density than Korean pine. In
the second layer there is some Korean pine, spruce, fir,
and linden. The growing stock is 400-500 m^/ha. The
dominant species on the forest floor are Carex calltnichos
and C. ussuriensis.

CoryluS'Carex-Korean Pine Forest— This type is
found on mountains with gentle slopes at altitudes of
700-900 m. As to favorability to tree growth, these sites
can be considered intermediate. Korean pine prevails
(60-70 percent) followed by Tilia amurensis (20 percent),
and some Abies nephrolepis, Picea jezoensis, Ulmus
propinqua, Acer mono, and Betula costata. The growing
stock is about 400-500 m^/ha. There are more tree species
associated, such as Acer sp. There is more undergrowth,
and the main species is Corylus mandshurica. On the for-
est floor about 30 species are to be found. Carex sidero-
sticta and ferns are the most common plants.

Fern-Spruce-Fir-Korean Pine Forest — This type
occurs on the lower gentle slopes (inclination about 10°)
of the mountains or along side streams at an altitude of
500-800 m. The community construction is complex. The
canopy layer is formed by Korean pine and some Populus
ussuriensis and Picea jezoensis. The second layer is com-
posed of Picea jezoensis and P. koraiensis (50 percent),
Abies nephrolepis (20 percent), and Tilia amurensis
(20 percent). Other species are Betula costata, Ulmus
laciniata, Acer mono, Phellodendron amurense, Juglans
mandshurica, Fraxinus mandshurica, and many kinds of
maple. Because of better site conditions, growing stock is
higher (250-350 m^/ha for the first story and 200 m^/ha for
the second story). The undergrowth is abundant with
over 20 species; about 30 species occur on the forest floor.
Ferns are dominant (such as Dryopteris).

Fern-Moss Korean Pine Forest — This type is found
in the lower part of mountains or in broad valleys; slope
inclination is 0-5°. Two stories can clearly be distin-
guished: Pinus koraiensis (80 percent) and a few Picea
jezoensis, Fraxinus mandshurica, Ulmus propinqua, Acer
mono, and Tilia amurensis form the forest canopy. In the
second story there is more Picea jezoensis, Betula costata,
Acer mono, Abies nephrolepis, Tilia amurensis, and a few
Juglans mandshurica, Fraxinus mandshurica, Ulmus
propinqua, Betula platyphylla, and many kinds of maple.

So the tree species are very abundant. This forest type
is typical of the sites most favorable to tree growth. The
growing stock can reach 600 m^/ha. The imdergrowth
mainly consists of Lonicera sp., Viburnum sp., and Sor-
baria, and is well developed. Carex and multiple fern and
luxuriant moss cover are common throughout the area.

COMMUNITY SUCCESSION
PATTERNS

Two succession patterns for the broad-leaved Korean
pine communities are obvious:

Succession Without Disturbance — Natural regen-
eration is poor under virgin Korean pine forest cover, al-
though a comparatively large quantity of seedlings may
appear after a seed year. However, they will only be able
to grow up if the canopy is cut thin. Because Korean pine
lives longer than other coniferous and broad-leaved trees,
and tolerates shade for longer period, it tends to form
multigeneration and multistoried forests.

Succession After Disturbance — ^After destruction of
broad-leaved Korean pine forest the succession goes on as
shown by the following sequences:

Selective cutting -» conifer-broad-leaved mixed forest
in which main elements are broad-leaved trees. Protec-
tion broadleaved-Korean pine forest.

Clear cutting or fire -> secondsiry bare land. Protec-
tion grass and shrub communities conifer-hardwood
mixed forests in which main elements are hardwood trees
-> conifer-broad-leaved mixed forest in which main ele-
ments are coniferous trees broad-leaved-Korean pine
forest.

To promote the reestablishment of Korean pine forest,
management should be adjusted to this succession
process.

CLIMATES WHERE STONE PINES
GROW, A COMPARISON

Tad Weaver

Abstract— While stone pine climates are similar, species
adapted to relatively moderate climates may be excluded from
the ranges of congeners by more severe climates, and species
with longer warm-moist growing seasons are probably more pro-
ductive than congeners. Absolute low/summer average/absolute
high temperatures for stone pines listed in order of increasing
absolute low temperature are Pinus sibirica (-55/13/37 °C),
P.pumila (-52/9/36 °C), P. koraiensis (-42/11/36 °C), P. albi-
caulis (-34/9/29 °C), and P. cembra (-23/8/27 °C). The Walter
drought index shows little stress in stone pine forests despite
large differences in summer/winter precipitation: in order of
increasing summer rainfall, precipitation is P. albicaulis (102/
829mm), P. pumiVa (142/264mm), P. sibirica {187 /245mm),
P. cembra (323/616mm), P. koraiensis (394/242mm). Estimated
thawed-soil growing season increases from P. albicaulis (4.5mo),
through. P. pumila (4.6mo), P. sibirica (5.5mo), and P. cembra
(6.3mo) to P. koraiensis (7.8mo); growing seasons of the first
three trees could be shortened by drought.

Stone pines grow in most of the boreal zone (fig.l)
(Fullard and Darby 1964; Lanner 1990; Mirov 1967).
Pinus pumila and P. sibirica occupy vast far-northern
(50 to 70° N. latitude) areas in northeastern Asia; P. sibir-
ica ranges from just west of the Urals (55° E. longitude)
halfway to the Pacific (115° E. longitude) and P. pumila
ranges from 115° E. longitude to the Pacific (165° E. longi-
tude). P. koraiensis grows at the boreal-deciduous forest
transition in eastern China (120 to 150° E. longitude and
45 to 55° N. latitude with outliers to 35° N. latitude at
high altitude). Paralleling P. koraiensis, P. albicaulis
occupies the Rocky-Cascade-Sierra Mountain chains of
North America from 55° N. latitude southward to 45° N.
latitude with outhers to near 35° N. latitude. Pinus
cemba appears in the Alps at approximately 45° N. lati-
tude. Glaciated parts of the conifer forest zone— both in
northern Europe and North America— lack stone pines
(fig.l).

Due to their common ancestry and common boreal
forest habitat, one expects the trees to occupy similar cli-
mates. This paper tests that hypothesis by comparing cli-
mates occupied by the trees with respect to 20factors that
may be important in determining the ranges and produc-
ivities of the species. The information may be especially
useful for predicting the success of introductions— from
one region to another— of stone pines, their tree associates,

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Statxis of Our Knowledge, St.Moritz,
Switzerland, September5-ll, 1992.

Tad Weaver is Plant Ecologist, Biology Department, Montana State
University, Bozeman, MT 59717.

and to some degree, even their herb, crj^jtogam, and ani-
mal associates. Such introductions might be of interest
as either producers or pests.

METHODS

Climates of environments dominated by closely related
stone pines (Lanner 1990; Mirov 1967) were compared.
The comparisons are nonstatistical, since the objective
was to discuss biological responses rather than the cli-
mates themselves.

The climates were characterized by choosing three to
four stone pine stands at which temperature and precipi-
tation were regularly measured (a complete sample of
accessible data); summarizing the data for periods of 10
years (P. albicaulis and P. cembra) or longer (unspecified);
and calculating means for selected parameters. The sta-
tions studied are Usted here— from west to east— each
with its approximate latitude, longitude, and altitude (m).
Note that P. albicaulis and P. cembra grow significantly
higher than, and to the south of, the Asian pines. Aster-
isked Siberian stations lack most temperature data. Sci-
entists who helped identify weather stations in stone pine
climates are listed in the acknowledgments.

Pinus albicaulis stations (USDC 1961-80) were Crater
Lake, OR (43° N., 122° W., 1,990m), Ellery Lake, CA
(38° N., 119° W., 2,940m), Old Glory, BC (49 ° N., 119° W.,
1,008m), and Kings Hill, MT (46 ° N., 110° W., 2,225m).
Pinus cembra stations (contributed by W.Tranquillini)
were Haggen (47° N., 11° E., 1,800m), Obergurgl (47 ° N.,
11° E., 2,070m), and Patscherkofel (48 ° N., 11° E.,
1,952m). Pinus sibirica stations (Muller 1982) were
Serov (60° N., 61° E., 132m*), Surgut (61 ° N., 73° E.,
40m), Kolpasevo (58 ° N., 83° E., 76m*), Jenisejsk
(58° N., 92° E., 78m*), Krasnojarsk (56 ° N., 93° E.,
15m), Tura (64 ° N., 100° E., 130m*), Irkutsk (52 ° N.,
104° E., 468m), and Kirensk (58 ° N., 108° E., 256m).
Pinus pumila stations (Muller 1982) were Vitujsk (64° N.,
122° E., 107m*), Jakutsk (62 ° N., 130° E., 100m*),
Verchojansk (68° N., 133° E., 137 m), Ochotsk (59° N.,
143° E., 6 m), Zyranka (65° N., 151° E., 43m*),
Petropavlovsk-Kamcatskij (53° N., 159° E., 32m), Apuka
(65° N., 170° W., 10m*), and Anadyr (65 ° N., 178° E.,
62m). Pinus koraiensis stations (contributed by Luo
Ju Chun and Zhao Shidong) were AnTu (43° N., 128° E.,
591m), Chargbei-ShenYang (42 ° N., 128° E., 738m) and
YiChun (48° N., 129° E., 231m).

Parameters studied describe climates with respect to
temperature, precipitation, and growing season. The fol-
lowing three paragraphs list these parameters and outline
the rationales for their choice; the rationales are devel-
oped further in Weaver (1993) and in the following
discussion.

7^ —

•1

Figure 1 — Distribution of stone
pines in relation to space, conifer
forests, and pleistocene glaciation.
North America and Eurasia are
seen from a polar view. Shading
locates the conifer zone on the two
continents (Fullard and Darby
1964). Areas within the dotted lines
received pleistocene glaciation
(Denton and Hughes 1981).
Ranges of the stone pines (Mirov
1967) are outlined by solid or
dashed lines: A = Pinus cembra,
B = P. sibirica, C = P. pumila,
D = P. koraiensis, E = P. albicaulis.
The absence of stone pines in the
conifer zone of northwestern
Eurasia and parts of North America
could be due to glaciation.

Ten temperature parameters were compared. Midwin-
ter frost danger was described by the long-term minimum
temperature (the absolute low), the average January
minimum, and the average January maximum. Extreme
fall and spring frosts are represented by the absolute low
in the first and last months of winter (that is, the absolute
lows in the months with OC average air temperature;
Weaver 1994). The average growing-season temperature
was calculated across those months when average air
temperatures were above 0 °C. This average was recalcu-
lated using "temperature growth support units = Q" to
account for the rise in rates of metabolic processes with
rising temperature; these points lie on a curve defined
by 0 °C = 0, 1 °C = 1, 11 °C = 2, 21 °C = 3, 31 °C = 4, and
41 °C = 5 (Weaver 1994). Summer highs were represented
by the average July minimum, average July maximum,
and long-term high.

Seven precipitation parameters were compared.
October- June precipitation was measured to determine
whether soil was moist in winter and at the opening of

the growing season. Summer precipitation was indexed
by July-September precipitation and by precipitation in
the wettest and driest month in that period. Since plants
require water, not precipitation, months with a positive
water balance were estimated with the Walter index
(Nielson 1986; Walter 1973), which assumes that for ev-
ery 2 °C rise in average temperature 1 mm of precipita-
tion will be evapotranspired. Two derivatives were stud-
ied: drought months (duration of periods with a negative
water balance) and drought magnitude (total estimated
deficit). While the Walter index tends to underestimate
drought (Stephenson 1990), it may be adequate in these
cool climates. Poikilohydric (Larcher 1975) organisms
may be more dependent on a number of growing hours set
by the number of rain days than on total precipitation;
rain days per month were therefore recorded.

Four growing-season parameters were compared. Length
of the growing season was indexed first as the number of
months when air temperature is above 0 °C, that is, the
approximate number of months when the soil is thawed

(Weaver 1994). Second, drought months (none according
to the Walter index) were subtracted from the warm-
season index to create a better warm-moist season index.
Despite contrary observations (Weaver 1994), growth is
expected to be indexed by the integral of temperature over
growing season. Thus two indices of growing season pro-
ductivity (W eaver 1994) were calculated: [growing season
X (average growing season temperature -5 °C) (Chang
1968)] and, considering the Q effect, [growing season
X average growing season Q].

WINTER TEMPERATURES

Average winter (January) temperatures in northeastern
Asia are the world's coldest and they warm southward
(Fullard and Darby 1964). Normal daily lows in regions
occupied by the trees are Pinus pumila (-30 °C), P. sibir-
ica (-27 °C), P. koraiensis (-27 °C), P. albicaulis (-14 °C),
and P. cembra (-8 °C), respectively (table 1). Daily highs
in winter average 8 °C higher (table 1).

Absolute lows experienced by Pinus sibirica (-55 °C)
and P. pumila (-52 °C) rise to P. koraiensis (-42 °C),
P. albicaulis (-34 °C), and P. cembra (-23 °C) (table 1).
Pinus albicaulis and P. cembra may be protected from ex-
treme low temperatures by the drainage of cold air from
their mountaintop sites. Becwar and Burke 1982 show
that timberline conifers of the Colorado Rockies do not
survive temperatures below -40 °C; thus P. sibirica and
P. pumila may be the only stone pines tolerant of -50 °C.

Frosts of early fall and late spring may catch trees in
partially hardened states. Absolute lows in the first and
last months of winter were -21 °C for P. koraiensis, -17 °C
for P. sibirica, and -10 °C for the remaining trees. Why
are absolute lows for fall-winter-spring frosts lower in the
P. koriaensis and P. sibirica regions than in the P. pumila
regions where average temperatures are lower (table 1;
Fullard and Darby 1964)?

SUMMER TEMPERATURES

Continentality allows northeastern Asia to warm in
summer more than less continental parts of the polar
region. Thus absolute maximum temperatures of Pinus
sibirica (37 °C) and P. pumila (33 °C) are considerably
higher than those of P. cembra (27 °C) and P. albicaulis
(29 °C) (table 1). Absolute summer highs in the P. korai-
ensis forest at the boreal-deciduous forest transition are
as high (36 °C) as in interior forests. Average July highs
in P. sibirica and P. koraiensis forests are like those in
Rocky Mountain Douglas-fir ( Pseudotsuga menziesii ) for-
ests, a vegetation zone below the Rocky Mountain stone
pine zone (Weaver 1994).

Daily growth rates are determined by average tempera-
ture conditions in the growing season. (1) Temperature
conditions are represented most simply by average tem-
perature; this is 13 °C for Pinus sibirica, 11 °C for P. korai-
ensis, and 8 to 9 °C for P. pumila , P. albicaulis, and P. cem-
bra. Like July highs, 13 °C average temperatures, are
similar to those found in the Douglas-fir/ponderosa pine
forests of the Rocky Mountains (Weaver 1994). (2) Be-
cause growth rises exponentially with increasing tem-
perature (Qjq; Larcher 1975), a better index of tempera-
ture on growth may be a cross-season average of growth
support units (Weaver 1994), which give greater weight
to high than low temperatures. With this index climates
cool from P. sibirica to P. albicaulis to P. pumila-P. korai-
ensis to P. cembra (table 1).

PRECIPITATION

Winter-spring precipitation is high in Pinus albicaulis
(829 mm) and P. cembra (616 mm) and lower (about
250 mm) in northeastern Asia (table 2). Since the lower
amounts will saturate most moimtain soils (Weaver 1978),
the excess is expected to run off, to have little effect on
water supplies in forest stands during the growing season.

Table 1 — Temperatures

^ in stone pine communities of the world.

Climates are listed in order of increasing winter temperatures

Species, location, number of stations

P . pumila

P. sibirica

P. koraiensis

P. albicaulis

P. cembra

Temperature

N. China

Siberia

Korea-China

N. America

Euro-Alps

data

Winter Temperature

Jan. mean min

-30.0 ± 8.0

-27.0 ± 2.0

-14.0 ±2.0

-8.0 ± 0.0

Jan. mean max

-24.0 ± 9.0

-19.0 ±2.0

-11.0 ±2.0

-5.0 ± 3.0

-1.0 ± 1.0

Abs min

-52.0 ± 4.0

-55.0 ± 2.0

-42.0 ± 1.0

-34.0 ± 2.0

-23.0 ± 1 .0

Frost spring

-11.0 ±3.0

-18.0 ±2.0

-26.4 ± 0.2

-1 1 .0 ± 1 .0

-10.0 ± 1.0

Frost fall

-11.0 ±3.0

-16.0 ±2.0

-27.6 ± 0.7

-1 1 .0 ± 1 .0

-10.0 ± 1.0

Summer Temperature

July mean min

8.0 ± 0.0

12.0 ± 1.0

14.0 ±1.0

4.0 ± 1.0

5.0 ± 1.0

July mean max

15.0 ±2.0

21.0± 1.0

26.0 ± 0.0

18.0 ± 1.0

14.0 ±1.0

Abs max

33.0 ±1.0

37.0 ± 1.0

36.0 ±2.0

29.0 ± 1.0

27.0 ± 2.0

9.0 ± 1.0

13.0 ±1.0

11.3± 1.2

9.0 ±1.0

8.0 ± 1.0

1.9 ±0.2

2.4 ± 0.1

1.9 ±0.1

2.1 ±0.2

1.7 ±0.1

'Temperature data (°C) are the mean ± one standard error. Absolute temperatures are recorded for 1 0 years in P. albicaulis and P. cembra;
records for the Asian pines are unspecified (presumed longer). T^, and Q^, are growing season averages defined in the text.

^ Sample size is four for P. albicaulis (except average max and min for January and July, n = 5), three for P. cembra, and for both P. pumila and
P. sibirica eight, except for January max-min, July max-min, and spring-fall frost temperatures where n = 4.

and, thus, to have httle effect on production. Large snow-
falls in the P. albicaulis and P. cembra forests, relative
to those in the Asian forests, surely result in greater
snowpacks, which shelter ground-level plants and ani-
mals less frost tolerant than the trees.

Summer precipitation is lower where moisture carrying
air masses cross mountains (northeastern Asia and the
Rocky Mountains) than where they do not (Alps and east-
ern China). Thus summer rainfall increases from Pinus
albicaulis (102 mm) to P. pumila (187 mm), P. sibirica
(187 mm), P. cembra (323 mm), and P. koraiensis
(394 mm). Precipitation in the driest month parallels
summer rainfall (table 2).

Soil water may be adequate in spite of low rainfall if
evapotranspiration is low. If we assume that 1 °C degree
evaporates 2 mm per month (Nielson 1986, 1992; Walter
1973), we see that stone pines do not experience drought
in any month (table 2). In addition, some buffering
against drought occurs because snowmelt water stored in
the soil provides a supplement to summer showers. How-
ever, even at boreal temperatures Walter's index may un-
derstate evaporation (Stephenson 1990), so there is prob-
ably less summertime restriction of growth by drought
(stomate closure) in Pinus cembra and P. koraiensis than
in P. albicaulis, P. sibirica, and P. pumila forests.

Organisms without water reserves — such as lichens,
mosses, and invertebrates — are more sensitive to summer
drought and may distinguish wetter and drier forests.
These organisms are surely inhibited in Pinus pumila
and P. albicaulis forests (one to eight rain days/month)
relative to P. cembra and P. koraiensis forests (16 to 18
rain days).

GROWING SEASON

Survival depends on photos3aithetic provisioning for
winter respiration and the outcome of competition may
depend on excesses above this basic provisioning. One
expects production to be correlated with the number of
warm-moist days, with the warmth of those days ( Q^^,
Larcher 1975), and thus with their product.

If water stress never occurs, the length of the warm-
moist season increases from Pinus albicaulis (4.5 mo)
to P. pumila (4.6 mo), P. sibirica (5.5 mo), P. cembra
(6.3 mo), and P. koraiensis (7.8 mo) (table 3). And if tem-
perature conditions were identical in these forests, one
would therefore expect production in P. cembra and
P. koraiensis forests to be half again as great as in
P. pumila and P. albicaulis forests. Actual differences
could be even larger because the temperature-defined
growing season may overstate stand productivity in drier
regions. For example, while subsoils remain moist all
summer in higher (J. Brown, personal commimication)
and lower (Weaver 1974) parts of the P. albicaulis zone,
drying of surface soils apparently causes tree water stress
at lower (B. Keane, personal communication), but not
higher (J. Brown, personal communication), sites. Similar
droughty periods probably occur in warmer P. pumila
and P. sibirica forests.

Production is expected to rise exponentially with rising
temperature, and thus productivity might be better corre-
lated with the product of season length and temperature
"growth supporting units." While this production hy-
pothesis was rejected in a cross-vegetation zone analysis
(Weaver 1994), the genetic similarity of stone pines might
allow it to operate here. If so, and if water deficits in
Siberia do not shorten the growing season, P. koraiensis
and P. sibirica will be promoted to the most productive
stone pines (table 3).

CONCLUSIONS

From a tree's point of view, average conditions in the
stone pine zones are similar: water stress is slight or
nonexistent and growing season temperatures average
9 to 13 °C. Seasonal extreme conditions may, however,
prevent reciprocal transplantation. Winter or fall-spring
frosts in the Pinus sibirica/P. pumila/P. koraiensis region
may exclude the other pines. And weak droughts in the
P. albicaulis/P. pumila/P. sibirica regions might exclude
P. cembra or P. koraiensis.

Table 2 — Precipitation^ in stone pine forests of ttie world. Climates are listed in order of increasing summer precipitation

Species, location, number of stations ^

P . albicaulis

P. pumila

P. s ib irica

P. cembra

P. koraiensis

Precipitation

N. Annerica

N. China

Siberia

Euro-Alps

Korea-China

data

Total

931 ± 229

407 ±137

432 ± 21

939 ± 9

636 ± 68

Oct.-June

829 ± 234

264 ± 76

245 ± 22

616 ±43

242 ± 62

July-Sept.

102 ± 14

143 ± 43

187 ±11

323 ± 36

394 ± 17

Wettest summer month

116± 16

165 ± 25

181 ± 8

214± 15

346

Driest summer month

4± 4

4± 2

8± 2

45 ±16

Summer drought months

0± 0

Summer water deficit

0± 0

Summer rain days,

average number

8± 1

1 ± 1

14± 1

16± 1

'Precipitation (mm) data are total (sum of all months), winter, summer (July, August, plus September), driest summer month (July-September) recorded,
wettest summer month (July-September) recorded, and average number of showers in June-September. High variances in total precipitation for P.
albicaulis and P. pumila are reduced to 705 ± 51 mm and 274 ± 37 mm by omission of the Crater Lake and Petropavlovsk stations, respectively.

^Sample size is four for P. albicaulis, three for P. cembra, three for P. Koraiensis, and eight for both P. pumila and P. sibirica, except for summer drought
months and deficit, where it is only four.

Table 3 — Stone pine productivity correlates: growing season ,^ growing season temperature,^ and their products. Climates are listed in
descending order by one estimate of productivity, warm moist season

Species, location, number of stations

P. koralensis P . cembra P. siblrica P. pumlla P. albicaulis

Korea-China Euro-Alps Siberia N. China N. America

Warm season

7.8 ±0.1

6.3 ± 0.2

5.5 ± 0.2

4.6 ± 0.4

4.5 ± 0.3

Summer drought months

0.0 ±0.0

0.0 ± 0.0

0.0 ±0.0

Warm-moist season

7.8 ±0.1

6.3 ± 0.2

5.5 ± 0.2

4.6 ± 0.5

4.5 ± 0.3

GS X Tg -5 °C
GSxQg,

11.3±1.2

8.0 ±1.0

13.0 ± 1.0

9.0 ± 1.0

1.9 ±0.1

1.7 ±0.1

2.4 ±0.1

1.9 ±0.2

2.1 ±0.2

56.5

18.9

44.0

18.4

18.0

15.6

10.7

13.2

8.7

9.5

'Warm-season and warm-moist season are two indices of growing season. Warm-season months occur after average monthly air temperature hses
above 0 °C and before it falls below 0 °C. The warm-moist season Is the warm season minus any months in which T/2 is greater than P(mm), Walter 1973;
"Walter drought" does not occur in stone pine regions.

H'wo indices of growing-season temperature are provided. T^, is the average temperature in growing season months. T -5C is used as one index of
growth support on the assumption that growth does not occur below 5 °C (Chang 1 968). Q is the average of temperatures weighted tor their growth-
supporting capacity: 0 °C = 0, 1 °C = 1 , 1 1 °C = 2, 21 °C = 4 {Weaver 1993).

Organisms filling other niches might find the climates
dissimilar (Weaver 1990). Poikilohydric organisms and
himaans would react differently to the relatively smnmer-
rainy climates of the P. cembra and P. koraiensis regions
than to the other stone pine climates. And organisms
wintering imder snow — small mammals, insects, plants
of small stature — are more protected from the extremes
of winter cold in the P. cembra and P. albicaulis regions
than in the other climates.

ACKNOWLEDGMENTS

I thank (1) W. Tranqmllini (University of Innsbruck,
Austria) for providing data from his P. cembra weather
stations, (2) J. Luo (Beijing Forestry University, China)
and Z. Shidong (Inst. Appl. Ecol., Shenyang, China) for
providing data from weather stations in P. koraiensis cli-
mates, (3) H. Lieth (University of Osabuck, Germany),
P. Gorchakovsky (Acad. Sci. Inst. Ecol., Ekaterinburg,
Russia) and P. Khomentovsky (Kamchatka Inst. Ecol.,
Petropavlovsk, Russia) for help in selecting P. sibirica
and P. pumila weather stations, and (4) nineteen North
American ecologists already acknowledged (Weaver 1990)
for help in selecting North American P. albicaulis stations.
I am also grateful for support to the project provided
through Intermountain Research Station contract #INT-
92720-RJVA (W. Schmidt).

REFERENCES

Becwar, M.; Burke, M. 1982. Winter hardiness limitations

and physiography of woody timberline flora. In: Li, P.;

Sakai, A. Plant cold hardiness and freezing stress. New

York: Academic Press: 307-324.
Chang, J. 1968. Climate and agriculture. Chicago: Aldine.

304 p.

Fullard, H.; Darby, H. 1964. The imiversity atlas. London:
George Philip and Son. 176 p.

Lanner, R. 1990. Biology, taxonomy, evolution, and ge-
ography of stone pines of the world. In: Schmidt, W.;
McDonald, K., comps. Proceedings — symposium on

whitebark pine ecosystems: ecology and management
of a high-mountain resource. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Research Station: 14-24.

Larcher, W. 1975. Physiological plant ecology. New York:
Springer. 252 p.

Mirov, N. 1967. The genus Pinus. New York: Ronald
Press. 602 p.

Muller, M. 1982. Selected climatic data for a global set
of standard stations for vegetation science. The Hague,
Netherlands: Junk. 306 p.

Neilson, R. 1986. High resolution climatic analysis and
southwest biogeography. Science. 232: 27-33.

Nielsen, R. 1992. Toward a rule based biome model. Land-
scape Ecology. 7: 27-43.

Stephenson, N. 1990. Climatic control of vegetation distri-
bution: the role of water balance. American Naturalist.
135: 649-670.

Walter, H. 1973. Vegetation of the earth in relation to cli-
mate and ecophysiolgical conditions. New York:
Springer. 237 p.

Weaver, T. 1974. Root distribution and soil water regimes
in nine habitat types of the northern Rocky Mountains.
In: Marshall, J., ed. The belowgroimd ecosystem. Sci.
Ser. 26. Fort ColHns, CO: Colorado State University,
Range Science Department. 351 p.

Weaver, T. 1979. Changes in soils along a vegetational
(altitudinal) gradient of the northern Rocky Mountains.
In: Youngberg, C, ed. Proceedings of the Fifth North
American forest soils conference. Madison, WI: Soil Sci-
ence Society America. 14-29.

Weaver, T. 1990. Climates of subalpine pine woodlands.
In: Schmidt, W.; McDonald, K., comps. Proceedings —
s)rmposium on whitebark pine ecosystems: ecology and
management of a high-moimtain resource. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
72-79.

Weaver, T. 1994. Vegetation distribution and production
in Rocky Mountain climates — wdth emphasis on
whitebark pine [these proceedings].

Growth Characteristics

International Workshop
St. Moritz 1 992

GROWTH OF SWISS STONE PINES
THAT ORIGINATED FROM AND WERE
PLANTED AT SEVERAL ALTITUDES IN
THE AUSTRIAN ALPS

Kurt Holzer

Abstract — For the purpose of gene conservation, a plus tree
collection of Swiss stone pine {Pinus cembra) was conducted.
Scions of 185 trees and 82 open-pollinated seed samples were
collected within the natural range of Swiss stone pine in Austria,
In 1987, clonal height was measured at an age between 25 and
30 years. Open-pollinated families were planted at three differ-
ent sites. Height was measured after 18 years. Growth of graft-
ings and open-pollinated families helped identify the altitude of
their origin. When planted below timberline, growth decreased
with increasing altitude.

In 1957, the Department of Forest Tree Breeding and
Genetics, Federal Forestry Research Institute, Austria,
commenced a selection program of Swiss stone pine (Pi-
nus cembra). The objectives of this program were to
preserve genetic resovirces of superior trees in a clonal
orchard and to assess the growth capacity of the clones
by field performance of open-pollinated families.

Within the natural range of stone pine, timberline di-
vides the distribution of stone pine, due to environmental
conditions, into two silvicultural zones: (1) production for-
ests and (2) "Kampfzone." Hence, those aspects had to be
considered for the program (Holzer 1963, 1976).

MATERIAL AND METHODS

Covering the majority of natural stands in Austria,
185 superior trees were selected between 1958 and 1965
(Holzer 1961, 1969). Within production forests scions of
124 and above timberline ("Kampfzone") scions of 61 indi-
viduals were collected. The altitudinal range was approxi-
mately 250 m below and 230 m above timberline (be-
tween 1,650 m and 2,200 m above sea level). In addition,
open-pollinated seeds were harvested from 82 trees.

Graftings were planted in a conservation orchard in
Purkersdorf-Stadlhutte near Vienna at 400 m above sea
level. Progenies were raised in the nursery of the insti-
tute (Mariabrvmn Vienna, 220 m above sea level) for
6 to 8 years and then planted at three different sites:
(1) Purkersdorf-Stadlhutte; (2) GroBe Zirbenwiese (1,650 m
above sea level); Seetaler Alpen, a swampy meadow
within the natiiral range of stone pine, approximately

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Kurt Holzer is Professor of Forest Genetics, Institut fur Waldbau,
Universitat fur Bodenkultur, A-1190 Vienna, Austria.

250 m below timberline; and (3) Melcheben, Packalpe
(1,700 m above sea level), close to timberline of Norway
spruce (Picea abies). At sites 2 and 3, only 32 families
were planted because of limited nvmabers of plants.

In 1987, height was measvired at a clonal age between
25 and 30 years. Heights of progenies were assessed at
an age of 8 and 18 years, respectively. Exclusively de-
scriptive statistics were employed.

Some results have been previously published with
respect to graftings (Holzer 1989) and progenies (Holzer
1978). Recently, growth of vegetative and generative
plant material has been measured (Feuersinger 1992).

Since altitude of timberline is irregular within the
Alps, comparisons are based on relative distances to local
timberline.

RESULTS AND DISCUSSION

Results are summarized in figures 1-3 within the main
distribution (about 150 m below to 100 m above timber-
line); mean height of 30-year-old graftings decreased with
increasing elevation of origin fi-om 3.8 m at 150 m below
to 2.7 m at 100 m above timberline. Variation was high
and yielded to 50 to 70 percent of the clonal mean. In
general, the majority of graftings originating below tim-
berline had superior growth. Presimiably, inherent
growth capacity shows a cline within the main distribu-
tion zone of stone pine (fig. 1).

This growth pattern was also found in generative plant
material. Progenies originating fi'om 100 m below timber-
line showed pronoimced reduced growth. However, when
growth was only evaluated within a single seed zone, it
was always higher below the timberline than above (fig. 2).
In the nursery, height of families (n = 50) originating fi'om
production forests averaged 46.5 cm. Families in = 16) fi'om
the "Kampfzone" had a slightly reduced growth of 40.0 cm
at an age of 8 years.

Transplants of families 18 years old showed different
results. Clinal growth pattern is still distinguishable at
site 1 (400 m above sea level) and site 2 (250 m below tim-
berline). However, at site 3 (timberline) height growth re-
actions were different. Here, families with small height
growth potential at 400 m above sea level were superior.
Figure 3 shows growth of families originating fi'om two
different seed zones. In seed zone "Defi-eggen" growth
of 18 families, 400 m and 1,700 m above sea level, was
negatively correlated (r = -0.752***). Fovuteen families
originating from seed zone "Prankerhohe" showed a con-
sistent pattern. However, correlation (r = -0.205) was not
significant.

200 -X

150-

100-

£ 50-

O above

f timber line
below

50 -
100-

150-
200-
250-

O X X J X X S

fix X

-|^T< 5-5-xW^X-X-^->

XXX X.J

X X Sx X S X X X >j6^ I i ^

XX Jgx X^xg)^ X

I \

X xx8xi>x'*8 X X

I X

X X I XX X

X S I X P XX X

I I I
3 4

Height of Graftings (m)

Figure 1 — Total height growth of 30-year-olcl
graftings In the plantation at 400 m above sea
level with the altitude of their origin (distance
from the local timberline). Dotted curve denotes
the mean values.

130

120

110 -

100

90 -

1 ,700 m asl

o o

X \

x"^ X

X \

400 m asl

100

120 140

Height (cm)

160

Figure 3 — The comparison of the mean of the
1 8-year-old families on two different planting
sites: 400 m and 1 ,700 m above sea level, re-
spectively (provenance 1/6/100, Prankerhohe,
marked with o, provenance 1/8/100, Defreggen,
marked with x; see text also). Curve denotes
correlation (r = -0.752) of the Defreggen zone.

This observation is supported by a sowing trial 200 m
above local timberline (2,100 m above sea level). Pro-
genies originating from sites above timberline performed
well; progenies from trees in production forests performed
poorly (Holzer 1975).

What conclusions regarding growth of stone pine might
be drawn from these experiments so far?

1. Height growth of stone pine shows clinal variation.

2. When plants are grown in production forests, growth
is positively correlated to elevation.

3. Above timberline this correlation is negative.

200-
150-
100-

•^50- e
c

S above
timber line
o below

•E 50-
<

100-

4Q 0

-X-XQW-^

o + ++ oma _pco>ooo 0 o o o

+ + + Ot +

150-

200-

■H-X |X X"xp X XX

-H-O-l- +■

50 60

Height of Progenies (cm)

Figure 2 — Total height of the 8-year-old prog-
enies (mean of families) of the harvested
clones with the altitude of their origin (distance
from the local timberline). Dotted curve de-
notes the mean values.

REFERENCES

Feuersinger, P. 1992. Das Wachstum von Zirbe in
Abhangigkeit vom Ursprungsort. Diplomarbeit, Univ. f.
Bodenkultur, Wien. 140 p.

Holzer, K. 1961. Der derzeitige Stand der Auswahl von
Schwarzerlen und von Zirbe im Rahmen der
Forstpflanzenziichtung. Informationsdienst 50, Allg.
Forstztg. 72.

Holzer, K. 1963. Pinus cembra L. as a pioneer at timber-
line in the European Alps. Proc. World. Cons. Forest
Genet, and Tree Imprv., Vol. 1 (FAO/FORGEN 1963,
3/13, FAO, Rome).

Holzer, K. 1969. Erste Ergebnisse von Zirbenauswahl-
baumen {Pinus cembra L.). Centrbl.f.d. ges. Forstwes.
86: 149-160.

Holzer, K. 1975. Genetics of Pinus cembra L. Annales

Forestales, Zagreb. 6(5): 139-158.
Holzer, K. 1976. Breeding haploxylon pines for subalpine

regions. XVI lUFRO-Congr. Oslo. Div. II: 216-227.
Holzer, K. 1978. Versuchsanlage GroBe Zirbenwiese. Exk.

Fiihrer Arbeitsgem. f. Hochlagenaufforstung. 5 p.

[Mimeo.]

Holzer, K. 1989. Drei Jahrzehnte Erfahrungen mit
Zirbenpfropfimgen. Centrbl. f.d. ges. Forstwes. 106:
79-88.

SEASONAL PATTERNS OF GROWTH
AND PHOTOSYNTHETIC ACTIVITY OF
PINUS PUMILA GROWING ON THE KISO
MOUNTAIN RANGE, CENTRAL JAPAN

Takuya Kajimoto

Abstract — Seasonal patterns of growth and photosynthesis in
Pinus pumila needles were investigated in relation to carbon
uptake. Current needles elongated from July to late August.
Photosynthetic activities of 2-year-old needles, both in sun- and
shade-needles, reached seasonal maxima in August, while those
of current needles did so in September. The results suggest that
annual net carbon gain of P. pumila depended mainly on net pho-
tosynthesis of old needles (1 to 4 year olds).

Growth and production rates of trees within the timber-
Une ecotone are generally restricted mainly due to the
short growing season (Tranquillini 1979). However, most
of the subalpine species show physiological adaptations to
conditions associated with low temperature, lower photo-
synthetic optimun temperature (Larcher 1975), and
higher frost and freezing resistance (Sakai and Larcher
1987). It has also been suggested that within krummholz
mats consisting of Picea engelmannii and Abies lasiocarpa
photosynthesis is enhanced by more favorable canopy-
temperature conditions (Hadley and Smith 1987). Infor-
mation on such physiological and structural functions is
indispensable for understanding how alpine trees survive
by achieving positive carbon balance during the short
siunmers.

In Japan, Pinus pumila Kegel, occurs in high-moimtain
areas from central Honshu (the main island) to the north-
em island of Hokkaido. This pine generally exhibits
dwarfed growth forms, and regenerates by layering at
the mature growth stage (Kajimoto 1992; Okitsu and Ito
1984). The physiognomy of P. pumila is similar to Pinus
mugo at the upper timberline in the European Alps or in
the Carpathian Mountains (Holtmeier 1973, 1981; Wardle
1977).

The foliage biomass of P. pumila rsinges from 15 to 24
tons d.w./ha despite the low tree heights. Pinus pumila
stands develop dense canopies with leaf area density of
about 5 mVm^, which is considerably more than in other
conifer forests (Kajimoto 1989a). This canopy structure
may result in large annual net production (Okitsu and Ito
1989; Shidei 1963). However, seasonal changes in growth

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Takuya Kajimoto is a special Researcher of Science and Technology,
Laboratory of Plant Production, Forestry and Forest Products Research
Institute, Tsukuba, Ibaraki 305, Japan.

and photosynthesis of P. pumila needles are still un-
known in relation to their carbon uptake.

In this paper, a process of photosynthetic production
in P. pumila during the growing season is discussed
based on observations of environmental factors and
seasonal patterns of growth and photosynthesis in the
needles.

STUDY AREA

The study area is located on the Kiso Mountain Range
in central Japan. The range trends from south to north.
The highest peak is Mount Kisokoma (2,965 m a.s.l.).
In this mountain range, P. pumila trees naturally occur
above 2,500 m. The study was carried out in the
Shinshu University Experimental Forest (35°48' N.,
137°50' E.) in the northern part of the Kiso Mountain
Range. The investigated pine stand is located on the
northeast slope (15° inclination) at an altitude of
2,600 m a.s.l. Dense canopy layer is developed between
100 and 200 cm above the ground height (Kajimoto
1989a).

FIELD OBSERVATIONS

Four siin-exposed terminal shoots of P. pumila were
selected. Lengths of current shoots, current needles,
and winter buds were measured seven times between
late May and October 1988. Current needle length of
each shoot was determined as the average of 10 fascicles
selected at equidistant intervals along the shoot axis.

At the plot, solar irradiance (400-1,100 nm in wave-
length) above the canopy siuface was measiu-ed with a
pyranometer sensor (LICOR, LI-200SB) at 1-hour inter-
vals from June to October 1988. Photos3nithetic photon
fliix density (PFD) (400-700 nm) above the canopy sur-
face and air temperature inside the stand were also
recorded at 1-hour intervals between mid-July and
October 1988 using a quantum sensor (Koito, IKS-25)
and a resistance thermometer (Koito, OPT-150).

MEASUREMENT OF
PHOTOSYNTHESIS

Rates of COg exchange were determined on detached
shoots that were collected 50 hours prior to the meas-
urement. The shoots were carefully transported to the
laboratory to minimize error associated with excision
and transportation (Kajimoto 1990). An open system

with an infrared gas analyzer (Hartmann and Braun,
URA-2) was used. The air was fed into an assimilation
chamber (9 by 25 by 1 cm^) at a rate of O.Sl/min. COg con-
centration of the input air was within the range of 360-
400 ppm. The chamber was placed in a thermoregulated
waterbath and illimiinated with incandescent lamps.
PFD at the chamber surface was measured with a quan-
tum sensor (LICOR, LI-190SB). Details of the system
were described by Kajimoto (1990).

Two terminal shoots (about 10 cm in length) of P.
pumila were sampled from the sun-exposed canopy sur-
face part (1.8-2.0 m) and in the lower shaded canopy lay-
er (1.0-1.2 m) seven times between late May and October
1988. At each sampling time, about 26 fascicles of ciirrent
needles (flushed in 1988) and 2-year-old needles (flushed
in 1986) were separately removed from each sun- and
shade-shoot. The needles were placed in the assimilation
chamber keeping the cut end of each fascicle immersed in
water using a small vinyl tube. Net photosynthetic rates
at different PFD levels (20-1,300 ^imol/mVs) and dark res-
piration rates of these two needle groups were measured.
Air temperature inside the chamber was maintained at
10 °C throughout the measurement; this temperature
value was within the range of optimum air temperatiire
(10-15 °C) in net photosynthesis for P. pumila needles
(Kajimoto 1990).

20 -

< 0

-5

-10
25

max.

c
o

20 -

£^ 5
o

^ 0

J A
1988

0 N

Figure 1 — Seasonal changes in daily solar irradiance
(400-1 ,100 nm in wavelength) and air temperatures
during 1988 measured in a P. pumila stand (2,600 m
a.s.l.), Kiso Mountain Range. Each point stands for
the mean value at 1 -week intervals.

■*-'
C
Ct

—I O

Figure 2 — Seasonal growth patterns of current
shoots, current needles, and winter buds ob-
served for sun-exposed P. pumila shoots. Each
point shows the mean for n = 4 shoots, o =
shoot length; • = needle length; □ = bud
length.

Needle area of each sample was measured as its pro-
jected area with an area meter (Hayashi, AAC-100), and
its dry weight was measured after ovendrying at 85 °C.
Specific leaf area (SLA; cm7g) was determined as the ra-
tio of needle area to dry weight. Net assimilation rates
obtained in all measurements were converted to the val-
ues at a COg concentration of 380 ppm.

RESULTS

Climatic Conditions

Daily solar radiation at the P. pumila stand was higher
in June and gradually decreased until October in 1988
(fig. 1). Mean air temperature was about 12 °C during
August and fell below 0 °C after mid-October. Monthly
precipitation was 212 mm in July and 167 mm in August
1988, according to the data obtained at Senjoujiki Station
(2,623 m a.s.l.) in the central part of Kiso Mountain Range
(Nagano Meteorological Observatory 1988).

Growth Patterns of Shoots and
Needles

The current shoots elongated rapidly from June to July
(fig. 2). The current needles began to flush in early July
and elongated rapidly during August. In late May, the
winter buds, which would flush in the next season, were
visible. The length of winter bud increased gradually un-
til late October.

Specific leaf area (SLA) of the current needles, both
sun- and shade-needles, was quite large in early August
when the needles began to grow larger, and sharply de-
creased by mid-September (fig. 3). SLA of 2-year-old

Cur -Shade

Figure 3 — Seasonal changes in specific leaf area
(SLA) of the P. pumila needles used for photosyn-
thesis measurements. □ = current, sun-needles;
■ = current, shade-needles; O = 2-year-old, sun-
needles;* = 2-year-old, shade-needles.

needles became slightly larger in August than in other
months. The shade-needles, both current and 2-year-old,
showed larger SLA values throughout the season than
sun-needles.

Cur.- Sun

Cur-Shod e
2 yr-Sun

2yr-Shade

1000 1200

Photosynthetic photon flux density
(pmol nn-2s~i )

Figure A — Examples of the relationships between
photosynthetic photon flux density (PFD) and net
photosynthetic rate (PJ at 10 °C of P. pumila
needles on September 15, 1988. □ = current, sun-
needles; ■ = current, shade-needles; 0= 2-year-old,
sun-needles; • = 2-year-old, shade-needles.

PFD

Light Response of Net Photosynthesis

Light response of photosynthesis was examined based
on the relationships between photon flux density (PFD)
and net photosynthetic rates (PJ at 10 °C (fig. 4). Light
saturation points of were different between sun- and
shade-needles, and also between needles of different age.
However, all sampled needles were completely light satu-
rated in at PFD value below 1,000 |imol/m7s and 10 °C.
A maximum net photos5nnthetic rate (P^^) was defined as
P^ value at 1,000 |imol/mVs and 10 °C, and an approxi-
mate light-saturation point in net photosynthesis (PFDg^)
was also determined as PFD value that gave 90 percent
ofP .

max

Table 1 shows that PFDgg of 2-year-old needles ranged
from 180 to 500 |imol/mVs in sun-needles and 180 to
440 |imol/m7s in shade-needles. The values were rela-
tively lower than those of the current needles, where
PFDgo was 440-580 [imol/mVs in sun-needles and 300-500
\imo]/mVs in shade-needles. Both current and 2-year-old
needles had highest PFDgg values in August compared to
the other months.

Seasonal Change in Photosynthetic
Activity

Figure 5 shows that P^^ of current needles were larger
in sun-needles than in shade-needles for each period. For
2-year-old needles, there was a remarkable difference in

between sun- and shade-needles in July and August.
The values of sun- and shade-needles both became higher
in August than in other months. P values of current

^ max

needles increased between late August and mid-September.

Dark respiration rates at 10 °C (R) of current needles
were larger in sun-needles than in shade-needles, while
there were no differences in 2-year-old sun- and shade-
needles (fig. 5). Current needles showed larger R values
than 2-year-old needles throughout the season.

Table 1 — Approximate light-saturation points (PFDgg) in the net
photosynthetic rates (at 1 0 °C) of P. pumila needles in
1988

Light-saturation points (PFD^Y

Date

Current needles

2-year-old-needles

Sun

Shade

Sun

Shade

May 27

420

440

July 2

340

July 20

300

August 4

500

380

August 24

580

460

480

380

September 1 5

440

360

180

October 20

500

300

240

220

'PFDgg was defined as the value of photosynthetic photon flux density
which gave 90 percent of a maximum net photosynthetic rate {P^^J- Each
PFDg^ value was determined using the light-photosynthetic curve, as shown in
figure 4.

«» 12

«A

* ft,

c o
w o 6

o e

<U _ £ ^
0

■4-^ MIC

a V 1.3

1 -
0.5-
0

/ ^ ^V-:- Cur-Sun

2yr-Sun

- 2yr-Shade

■N 0

Cur-Shade
— ' ' ' '

M J

J A S 0

2yr-Shade

2yr-Sun

1988

Figure 5 — Seasonal changes in net photosyn-
thetic rates {P^J at 1 ,000 iimol/m^/s and
10 °C, and dark respiration rates (fl) at 10 °C
of P. pumila needles during 1988. □ = current,
sun-needles; ■ = current, shade-needles;
O = 2-year-old, sun-needles; • = 2-year-old,
shade-needles.

Needle longevity of P. pumila, 4 years or more, corre-
sponds to that of P. mugo and P. contorta (Benecke and
Havranek 1980).

Climate Factors Affecting
Photosynthetic Production

Seasonal differences in photosynthetic activity between
new and old needles have been reported for some north-
em or subalpine evergreen conifers (Fry and Phillips
1977; Teskey and others 1984). Likewise, photos5Tithetic
activity (P^^J in the 2-year-old needles of P. pumila be-
came higher between late July and early August, while
that of current needles did so in late August and mid-
September (fig. 5). A similar seasonal trend in P^^ was
found for the current and older needles (1 to 4 years old)
of P. pumila in the study plot in 1987 (Kajimoto 1990).
These results suggest that photosynthetic production of
the P. pumila canopy is primarily active during August
and September.

Alpine conifer trees rarely achieve their potential maxi-
mum photosynthetic activities under natural environmen-
tal conditions, although the main limiting factors in photo-
sythesis, such as irradiance and air and soil temperatures,
seem to be different depending on species and site condi-
tions (De Lucia and Smith 1987; Hasler 1982; Tiirner and
others 1983). Mean air temperatures between late July
and mid-September at the research plot (fig. 1) corre-
sponded to optimiun air temperatm-e (10-15 °C) for net
photosynthesis of P. pumila, both for current and old

DISCUSSION

Growth Characteristics of Needles

Needles of P. pumila began to elongate in early July
and almost finished growth in late August (fig. 2). The
period of needle growth is 1 or 2 months later than those
of other low-altitude pines, for example, P. densiflora
(Tanaka and others 1976) and P. thunbergii (Nagatsu
1987), growing naturally in central Japan. After needle
growth had terminated, the current needles of P. pumila
showed gradual decline in specific leaf area (SLA) (fig. 3)
and maintained higher dark respiration rates than
2-year-old needles (fig. 5). This indicates that the current
needles continued to accumulate internal carbohydrate or
nutrient until October.

In the study area, there was little local variation in sea-
sonal growth patterns in the shoots and needles of P.
pumila (Kajimoto 1989b). However, the final lengths of
current shoots and needles became smaller with increas-
ing altitude, as reported for other high-altitude pines,
such as P. cembra (Baig and Tranquillini 1976), and P.
mugo and P. contorta (Benecke and Havranek 1980).

The current needles of P. pumila in the research plot
accounted for 9 percent of total foliage biomass in late
July and 27 percent in late August (Kajimoto 1989a).
This indicates that 70 percent or more of the pine canopy
consisted of older needles throughout the growing season.

^ August

^ 30-

u
c

Of
D
CT
<1>

"a
cr

20
10

^ '0

September

30
20
10
0

Air temperature CO

Figure 6 — Relative frequency distributions of
daytime (8 a.m. to 4 p.m.) mean air tempera-
ture at each 1-hour interval in August (above)
and September (below) 1 988. Air tempera-
tures were recorded at the P. pumila stand
(2,600 m a.s.l.), Kiso Mountain Range.

30-

o
c

u
cr

20 -
10 -

August

1000

2000

5 301-
(T 20h

September

10 ^

0 *^^^^°"'f"

0 1000 2000

PFD ( jjnnol m'h'' )

Figure 7 — Relative frequency distributions of
daytime (8 a.m. to 4 p.m.) photosyntiietic photon
flux density (PFD) at each 1 -hour interval in
August (above) and September (below) 1 988.
PFD was measured above the canopy surface of
P. pumila stand (2,600 m a.s.l.), Kiso Mountain
Range.

needles (Kajimoto 1990). Figure 6 shows patterns of rela-
tive frequency distribution of daytime mean air tempera-
ture (8 a.m. to 4 p.m.) at 1-hour intervals; temperature
values between 10 and 16 °C made up about 70 percent of
the sum total in both August and September. With re-
gard to irradiance conditions, proportions of PFD values
above 600 |j.mol/m7s, when the pine needles were photo-
synthetically light saturated (table 1), were 53 percent in
August and 36 percent in September (fig. 7). The tem-
perature and light regimes indicate that net assimilation
of P. pumila needles under natural conditions is limited
mainly by lower levels of solar irradiance. Reduction in
available PFD from August (53 percent) to September
(36 percent) is likely to impede net photosynthesis of the
current needles, since photosynthetic activities (P^^) of
the current needles were the highest in September (fig. 5).

It is concluded from the present study that P. pumila
needles are not very efficient in photosynthesis diiring the
first growing season. The pine needles survive for four or
more growing seasons and recover their photosynthetic
abilities in each summer; P of the oldest needles

' max

(4 years old) was about half of that of 1-year-old needles,
although P^^ decreased with needle age (Kajimoto 1990).
Consequently, a large proportion of annual net carbon
gain in the P. pumila tree is likely to depend on net as-
similation by the old needles.

Knowledge of physiological properties of P. pumila,
especially water relations (Ando and Kawasaki 1991), is
still limited compared to other subalpine pine species. As
to seasonal photosynthetic production of P. pumila, more
information on its photosynthetic reponses to environ-
mental factors, particularly based on field measurement,
is needed.

ACKNOWLEDGMENTS

I thank Dr. K. Hozumi for his suggestions and en-
couragement during this study, and Drs. T. Koike,
M. Ishizuka, A. Osawa, Y. Chiba, and N. Kurachi for their
useful comments on this manuscript. Comments by Drs.
F.-K. Holtmeier, W. C. Schmidt, and R. Hasler are ac-
knowledged. Thanks also go to Professors T. Horiuchi
and S. Ito, Shinshu University, and to the staff of Shinshu
University Experimental Forest and of Nagoya University
Forest for their assistance during the field siu"vey.

REFERENCES

Ando, H.; Kawasaki, K. 1991. Growth of Pinus pumila in
relation to the site conditions on a slope. Transactions
of the 102d Meeting, Japanese Forestry Society:
451-452. [In Japanese].

Baig, M. N.; Tranquillini, W. 1976. Studies on upper tim-
ber line: morphology and anatomy of Norway spruce
(Picea abies) and stone pine (Pinus cembra) needles
from various habitat conditions. Canadian Journal of
Botany. 54: 1622-1632.

Benecke, V.; Havranek, W. M. 1980. Phenological growth
characteristics of trees with increasing altitude,
Craigieburn Range, New Zealand. In: Benecke, U.;
Davis, M. R., eds. Mountain environments and subal-
pine tree growth. Tech. Pap. 70. New Zealand Forest
Service, Forest Research Institute: 155-174.

De Lucia, E. H.; Smith, W. K. 1987. Air and soil tempera-
ture limitations on photosynthesis in Engelmann
spruce during summer. Canadian Journal of Forest
Research. 17: 527-533.

Fry, D. J.; Phillips, I. D. J. 1977. Photosynthesis of coni-
fers in relation to annual growth cycles and dry matter
production. II. Seasonal photosynthetic capacity and
mesophyll ultrastructure in Abies grandis, Picea
sitchensis, Tsuga heterophylla and Larix leptolepis in
S.W. England. Physiologia Plantarum. 40: 300-306.

Hadley, J. L.; Smith, W. K. 1987. Influence of krummholz
mat microclimate on needle physiology and survival.
Oecologia. 73: 82-90.

Hasler, R. 1982. Net photosjTithesis and transpiration of
Pinus montana on east and north facing slopes at al-
pine timberline. Oecologia. 54: 14-22.

Holtmeier, F.-K. 1973. Geoecological aspects of timber-
lines in northern and central Europe. Arctic and Alpine
Research. 5(3): 45-54.

Holtmeier, F.-K. 1981. What does the term "krummholz"
really mean? Observations with special reference to
the Alps and the Colorado Front Range. Mountain
Research and Development. 1(3/4): 253-260.

Kajimoto, T. 1989a. Aboveground biomass and litterfall of
Pinus pumila scrubs growing on the Kiso Mountain
Range in central Japan. Ecological Research. 4: 55-69.

Kajimoto, T. 1989b. Current growth of Pinus pumila.
Transactions of the 37th Meeting, Chubu Branch,
Japanese Forestry Society: 55-58. [In Japanese].

Kajimoto, T. 1990. Photosynthesis and respiration of
Pinus pumila needles in relation to needle age and
season. Ecological Research. 5: 333-340.

Kajimoto, T. 1992. Dynamics and dry matter production
of belowground woody organs of Pinus pumila trees
growing on the Kiso Mountain Range in central Japan.
Ecological Research. 7: 333-339.

Larcher, W. 1975. Physiological plant ecology. Heidelberg:
Springer. 252 p.

Nagano Meteorological Observatory. 1988. Annual report
on weather conditions of Nagano prefecture in 1987.
Nagano Branch Office of Japanese Meteorological Soci-
ety. 137 p. [In Japanese].

Nagatsu, M. 1987. Correlation between the growth of
needles and shoots in a Pinus thunbergii yoimg tree.
Transactions of the 98th Meeting Japanese Forestry
Society: 417-418. [In Japanese].

Okitsu, S.; Ito, K. 1984. Vegetation dynamics of the Sibe-
rian dwarf pine {Pinus pumila Regel) in the Taisetsu
Mountain Range, Hokkaido, Japan. Vegetatio. 58:
105-113.

Okitsu, S.; Ito, K. 1989. Conditions for the development of

the Pinus pumila zone of Hokkaido, northern Japan.

Vegetatio. 84: 127-132.
Sakai, A.; Larcher, W. 1987. Frost survival of plants. Ecol.

Stud. 62. New York: Springer. 321 p.
Shidei, T. 1963. Productivity of Haimatsu {Pinus pumila)

community growing in alpine zone of Tateyam-Range.

Journal of the Japanese Forestry Society. 45: 169-173.
[In Japanese with English simimary].

Tanaka, H.; Oohata, S.; Akai, T. 1976. Elongation and
shoot forms in foreign pines. Report, Kyoto University
of Forestry. 11: 38-49. [In Japanese].

Teskey, R. O.; Grier, C. C; Hinckley, T. M. 1984. Changes
in photosynthesis and water relations with age and sea-
son in Abies amabilis. Canadian Journal of Forest
Research. 14: 77-84.

Tranquillini, W. 1979. Physiological ecology of the alpine
timberline. Ecol. Stud. 31. New York: Springer. 131 p.

Turner, H.; Hasler, R.; Schonenberger, W. 1983. Contrast-
ing microenvironments and their effects on carbon up-
take and allocation by young conifers near alpine
treeline in Switzerland. In: Wareing, R. H., ed. Carbon
uptake and allocation in subalpine ecosystems as a key
to management: Proceedings of lUFRO workshop. 1982
August 2-3. Corvallis, OR: Oregon State University,
Forest Research Laboratory: 22-30.

Wardle, P. 1977. Japanese timberlines and some geo-
graphic comparisons. Arctic and Alpine Research. 9:
249-258.

HEIGHT GROWTH IN CEMBRAN PINE
AS A FACTOR OF AIR TEMPERATURE

Herbert Kronfuss

Abstract — The study treats height growth in cembran pine
iPinus cembra) in high-elevation afforestation at an altitude
of 1,800 m with growth patterns analyzed as a factor of air tem-
perature. It is assumed that growth behavior follows an endog-
enous rhythm possibly deriving from the genotj^je. The relation-
ship between growth pattern and air temperature is demon-
strated for defined periods and phases of height growth. From
the results it becomes obvious that height growth and thus the
relative rate of growth are primarily dependent on temperature.
Data from long-term monitoring series were used to determine
the ampUtudes between the growth curves for early and for late
culmination. A comparison of the ampUtudes for cembran pine
with spruce and larch shows cembran pine — with the highest
amplitudes — to be the most sensitive to temperature. Cembran
pine thus makes most efficient use of warm weather periods for
increment production in short periods of time.

The cembran pine (Pinus cembra) occupies the highest
stands of all trees in the Alps and as such is ideally suited
for a study of height growth related to air temperature.
The fact that cembran pine is not thought to have differ-
ent ecotypes adapted to various altitudes also permits
growth characteristics of this species to be defined in spe-
cific terms.

STUDY SITE

The study site was located 30 km southwest of Innsbruck,
Tyrol (Austria), in the Stubai Alps near Haggen in the
Sellrain Valley. For location, geology, and climate details,
see figures 1 and 2, and table 1. The results relate to a
continuous experimental stand of cembran pine at an alti-
tude of 1,800 m. The stand was established in 1970 when
the plot was afforested with 4-year-old plants.

MONITORING

Monitoring was performed over a period of 8 years
for a group of 40 trees. Height growth in cembran pine
was measured at 5-day intervals (pentades) and the rate
of growth correlated with mean air temperature of the
pentade. Individual height increments were expressed
as a percentage of annual cumvdative grov^^h to permit
the growth patterns to be compared -with, changes in
temperatxire.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Herbert Kronfuss is Forester, Forestry Research Institute of Austria,
Rennweg 1/29, A-6020 Innsbruck.

Growth patterns are represented in the form of relative
growth-rate c;irves and relative cumulative-growth curves.
This not only permits direct comparisons to be made be-
tween annual increments but also, given a long enough
monitoring period, permits a characteristic increment pat-
tern (type) to be derived for each species (Burger 1926;
Larcher 1980; Week 1955).

HEIGHT GROWTH

From 1978 to 1985 increment growth lasted an average
of 108 days, vnth a 20-day variance range.

To study the influence of temperature on the growth
pattern during the period of extension growth, the rela-
tive grovrth-rate curve was correlated with the tempera-
ture curve, with the mean height increments from con-
secutive pentades expressed as a percentage of total
height growth. The relative growth-rate curve clearly
illustrates the pattern of height grov^h (fig. 3).

Figure 3 shows how temperature fluctuations during
the period of extension growth influence the rate of height
growth. With regard to the correlation between the
height-growth behavior and temperattire patterns, the
following points would seem to be of interest.

Dming growing season, the temperature curve itself
is characterized by two pronotmced dips for the periods
May 20 to 25 and Jtme 15 to 20, a phenomenon well
knovm to the local rural communities as the "Ice Saints"
and "Sheep's Chill." At the beginning of April, mean tem-
perattire reaches freezing point on the long-term average.
The rise in temperature continues imtil mid-May, reach-
ing an average of approximately 7 °C, before dropping
back to 5 °C for the pentade occupying the first of the two

Figure 1 — Geographical location of the site.

HAGGEN i.S.dSOO m) a) 3,1° b) 920 mm
c) 10 d) 1975 - 1984

mm
200

100

I I I I I

I I I I I I I

I I I I I

a) Mean annual temperature

b) Mean annual precipitation

c) Number of years of data collection for temperature and
precipitation

d) Period of data collection for temperature and
precipitation: 1975-1984

e) Mean monthly precipitation

f) Mean monthly temperature

g) Months with average daily minimum below 0°C

h) Months with absolute minimum temperatures below 0°C

Figure 2 — Data for Haggen Climate Diagram
Haggen I.S., Field Station 1, 1,800 m, SSW.

cold spells. During the time of the "Ice Saints," most of
the cembran pines are at the beginning of the period of ex-
tension growth, with a growth rate of only 2 to 3 percent
of relative growth per pentade. That corresponds to about
1.5 cm. On the average, approximately 7 percent of cumu-
lative growth is achieved by this point in time. For height
grovi^h in cembran pines in this location, the period be-
tween the "Ice Saints" and "Sheep's Chill" is of special
importance as it includes the first major peak in the tem-
perature curve, on or around May 31 on average.

With a brief delay, this temperature peak stimulates
a response in the cembran pine in terms of height growth,
and after one- and one-half pentades, the biggest absolute
increment is produced, accounting for about 35 percent of

total growth on a long-term average. Following this peak
in increment production, the typical pattern of height
growth in the cembran pine is one of declining rates of
growth in spite of rising temperatures. This is shown in
figure 3 with the two curves moving in opposite directions
following the start of the "Sheep's Chill" period aroimd
June 20.

Not even an increase in mean pentade temperature
beyond the 10 °C mark around July 5, finally peaking
at 13 °C, leads to a further spontaneous spate of growth.

Table 1 — Location of the study: HAGGEN afforestation site in the
Sellrain Valley near St. Sigmund, Tyrol

Geographical coordinates:
Height above sea-level:
Exposition and angle of slope:

47°13'N. ir06'E.
1,715-1,950 m
SSW, 30°

Bedrock: Foliated gneiss (biotite granite gneiss, granodiorite
gneiss) and mica schist.

Vegetation at

afforestation: Heath (callunetum, nardetum alpigenum)
Soils: Humus horizons developed into a brown earth type

through grazing and haymaking. More or less

podsolic brown soils.
In previous centuries a heavily grazed avalanche slope.

CLIMATIC DATA are based on a 10-year monitoring period (1975-
1984)

- Sunshine hours during the vegetation period (mid-May to

September)

a) theoretical asatronomical maximum 2,005 hours

b) theoretical local maximum 1 ,393 hours
horizon loss 30.5 percent

- Air temperature

a) mean annual temperature 3.1 °C

b) during vegetation period 8.4 °C

- Precipitation

a) mean annual precipitation (3 m aboveground) 920 mm

b) during vegetation period (3 m aboveground) 533 mm

c) precipitation at ground level 650 mm
(= approx. 22 percent higher than at

3 m aboveground)

- Precipitation probability (percent)

a) June maximum 55 percent

b) December minimum 27 percent

- Mean precipitation distribution (mm)

a) maximum in July and September 8.4 mm

b) minimum in February 3.6 mm

c) month with highest precipitation (July) 141 mm

d) month with lowest precipitation (February) 33 mm

- Evaporation during the vegetation period

a) potential evaporation (2.5 m aboveground 566 mm
with "piche") (= 4.1 l/m^ per day)

b) actual evapotranspiration (lysimeter with 236 mm
ground vegetation (= 1 .7 l/m^ per day)

c) actual hydrological budget 412 mm
(= 3 l/m^ per day)

The meteorological summer semester accounts for 65 percent of
annual precipitation (598 mm), and the meteorological winter
semester 35 percent (322 mm).

100

Mean pentad values for relative growth rate
and air temperature (°C)

Mean relative growth rate (%)

January February March April May June July August September October November December

Figure 3 — Comparison between the curves of relative growth rate and air
temperature, 8-year series (1978 to 1985).

The minor fluctuations in the growth curve on the order
of 1 to 3 percent can be correlated with temperature fluc-
tuations of 10 to 13 °C.

The main features of the curves in figure 3 can be de-
scribed as follows: The grovvi;h curve for cembran pine
peaks (maximum growth rate per pentade is reached)
on June 5 to 10 on average. The temperature curves are
characterized by two peaks, the first between the two pro-
nounced dips (May 31 to June 5) and the second between
July 25 and July 31. It is interesting to note that height
increments between July 5 and the end of the period of
extension growth — and that is a good 46 days — total
only 20 percent of cumulative grov\i;h, even though the
temperature curve passes 10 °C on July 5 and remains
more or less at that level until mid-September.

The above picture, based on the mean figures calculated
from monitored data, illustrates in bold strokes the corre-
lation between the temperature and height-growth patterns.

The relative grov^i:h-rate curves shown in figure 4 for
the 8 years from 1978 to 1985 provide an overview of
growth patterns in cembran pine as a factor of air tem-
perature for the individual years. One of the more strik-
ing features of these curves is the variation in the onset
of culmination (equals greatest increment production per
pentade). On this basis we can distinguish between an-
nual growth curves with early culmination (1979, 1981,
1982, 1983), one year with late culmination (1984), and
years with twin or intermediate peaks (1978, 1980, 1985;
see also table 2).

A comparison of the temperature curve in figure 3 with.
the growth-rate curves in figure 4 reveals two clearly
delineated time periods in which culmination (peak
growth rate per pentade) occurs. The first is the period
between the "Ice Saints" and "Sheep's Chill," and the sec-
ond is a subsequent period of 30 days ending approxi-
mately with the pentade of July 15 to 20. These two peri-
ods together total 57 days (equals 53 percent of the mean

1978

...z

■ 1

. . . . 1

!^^^7^ ^

10 20 31 10 20 30 10 20 31 10 20 31

May June July August

W 20 31 10 20 30 10 20 31 10 20 31

— »«

■

^^~.yrr>^

10 20 31 10 20 30 10 20 31 10 20 31

1980

■ ■1

W 20 31 10 20 30 10 20 31 10 20 31

1981

■A^

« 20 31 10 20 30 10 20 31 10 20 31

1983

,,l

10 20 31 10 20 30 10 20 31 10 20 31

1986

10 20 31 10 20 30 10 20 31 10 20 31

Figure 4 — Cembran pine relative height-growth
curves (percent).

101

Table 2 — Air temperature in °C and percentage growth in cembran
pine'

J n I I

Year

°C-

- - Percent - -

Early culmination

1979

245

9.1

298

9.9

1981

242

8.9

230

7.7

1982

246

9.1

368

12.3

O IT

1983

203

7.5

359

12.0

Average

234

8.7

314

10.5

Intermediate culmination

8.2

241

8.0

1980

195

7.2

205

6.8

1985

191

7.1

284

9.5

Average

203

7.5

243

8.1

Late culmination

1984

169

6.3

284

9.5

'I = period from the "Ice Saints" to "Sheep's Chill"; II = period from "Sheep's
Chill" to 20 July; TS = temperature sum; T = mean temperature; HG = height
growth; period I lasts for 27 days; period II lasts for 30 days.

10 20 31 10 20 30 10 20 31 10 20 31

May June July August

Figure 5 — Cembran pine relative height growth
(percent).

period of extension growth) and account for an average
of 82 percent of annual shoot length.

There is a significant correlation between percentage
growth within the two main growth periods and tempera-
ture sum, v^dth peak growth located in the period that ex-
hibits the highest temperature sum in absolute terms,

except that beyond an average temperature sum of 230 °C
for the first period a further increase in temperature for
the second period does not trigger higher growth rates in
that period. The highest pentade growth rate always oc-
curs in the period with the highest temperature Bum and
coincides with or follows the pentade with maximum
pentade temperature (see table 2).

From table 2 it can be seen that in the case of early cul-
mination the average temperature sum in the first period
was 234 °C (8.7 °C/27 days). During that period, an aver-
age of 58 percent of extension growth was measured (with
scatter between 46 percent and 71 percent). In those
cases the cembran pines achieved an average of only 28
percent relative growth in the second period in spite of the
rise in the average temperature sum to 314 °C (10.5 °C/30
days).

The year of late culmination, 1984, is also instructive.
The cold weather in period I of that year produced a tem-
perature sum of only 169 °C (6.3 °C/27 days), which is
reflected in a correspondingly low relative growth rate
of 17 percent. Period II, however, was characterized
by a significant rise in temperature, producing a tem-
perature sum of 284 °C (9.5 °C/30 days). The relative
growth rate in that period (fig. 5) was 58 percent, which
is identical with the figure for period I in the years of
early culmination.

At the end of July in the late culmination year,
the leaders had reached 90 percent of total extension
growth compEired with 98 percent for the years of early
culmination.

This analysis of the height-growth patterns of cembran
pine over an 8-year period of observation (1978 to 1985)
shows that the cembran pine is ontogenetically equipped
to make efficient use of available warmth for increment
production so as to achieve ctdmination as early as pos-
sible. Up to that point, air temperature is the dominant
factor in determining the relative growth rate, £ind after-
ward it continues to function as the "engine" that main-
tains the extension process.

PHASES OF HEIGHT GROWTH

The following analysis of annual height growth by help
of the growth curve (Kronfuss 1985) will illustrate the
somewhat abstract correlation between growth rate and
temperature with the aid of further figures. For this pur-
pose, height growth can be divided into the following
three phases (fig. 6):

a. Initial phase (0 to 25 percent of annual extension).

b. Main phase (25 to 75 percent of annued extension).

c. Final phase (75 to 100 percent of annual extension).

Table 3 provides a separate assessment of the correla-
tion for each of these three phases using a regression line,
correlation coefficient, and regression coefficient to ex-
press the relationship between mean daily temperature
(T/d) and average height growth per day (HG/d) (fig. 6,
table 3).

Whereas the correlation coefficient for the initial phase
(r = 0.6) suggests a correlation vnth temperature, this is
no longer the case in the final phase (r = 0.42). The main

102

Correlation between mean daily air temperature and height increment
per day HAGGEN i.S., 1800 m

Average height increment per day (mm)

r = 0.89 k=^^!£A

^ Mam phase

Initial phase

r=.0.42 /f.O.II

•

• •

r=o.60 ;f=o.is ■

• ••

1 i_

* " * Final phase

_ 1 ■

S 6 7 8

9 10

It 12 13 14

Air temperature ("C)

Correlation between mean daily air temperature and growth processes
HAGGEN i.S.. 1800 m

Duration of phases (days)

Air temperature (°C)

Figure 6 — Cembran pine correlations between
mean daily air temperatures, height increment, and
duration of growth phases: period 1 978 to 1 985.

growth phase, on the other hand, is largely determined by
temperature, as can be seen from the almost stretched
curve (fig. 6 at top) and a correlation coefficient of r = 0.89.

The regression coefficients {h), as the expression of aver-
age daily growth per vinit of temperature, show that, for
the same temperature rise, height growth in the main

phase is 4.5 times greater than in the initial phase and
6.2 times greater than in the final phase. In the initial
phase, with an average period of extension growth of 32
days, the average daily increment is 1.6 mm. In the
main growth phase the period of extension growth is re-
duced to an average of 26 days of an HG/d of 4 mm, while
the figure for the subsequent 50 days of the final phase is
only 1 mm.

The duration of extension growth also correlates with
temperature and is shorter with higher temperature rise.
The influence of temperature on the period of extension
growth can be seen from the correlation coefficients, with
a close correlation with temperature expressed by the co-
efficients r = 0.82 for the initial phase and r = 0.84 for
the main growth period, decreasing in the final phase
to r = 0.59. This low correlation coefficient shows that
reduced height growth in the final phase is no longer due
exclusively to the influence of temperature, but to the
tailing away of an endogenous pattern. With regard to
the temperature-dominated phases of extension growth,
(the initial and main phases) an increase in mean tem-
perature by 1 °C reduces the extension growth period by
4 to 5 days. Also, in the final phase we see again that the
pattern derives from an autonomous endogenous rhythm
that determines the potential period of extension growth
in this phase for the species involved.

GROWTH CURVE VARIANTS

As can be seen from the shapes of the relative growth-
rate curves in figure 4, annual patterns of height growth
vary considerably for the same species in the same loca-
tions. Eight years of observation offer enough data for
graphic representation of maximum amplitudes between
the growth curves for the earliest and those for the latest
culminations and for a comparison with spruce and larch
(fig. 7). This illustrates not only the time sequences with
regard to flushing but also the cxirrent range of fluctua-
tion for these tree species.

Table 3 — Correlations between mean daily temperature (t7d) and height growth per day (HG/d) and between mean dally
temperature (t7d) and the period of extension growth days (d)

Year

Initial phase^

Main phase^

Final phase^

t°/d

HG/d

t°/d

HG/d

t°/d

HG/d

1978

7.9

1.8

6.9

2.6

10.4

0.9

1979

9.6

1.8

8.8

4.7

9.9

1989

5.4

6.6

2.5

12.0

1.0

1981

6.5

1.7

4.7

8.9

1982

8.2

1.6

9.3

4.1

11.7

1983

6.1

1.8

8.8

4.6

13.1

1.5

1984

5.2

1.3

9.7

5.5

9.8

1.6

1985

7.5

2.1

7.2

3.5

11.2

1.4

Average

7.1

1.6

8.5

4.0

10.9

1.1

'Initial phase: correlation coefficient r= 0.604; regression coefficient /( = 0.147; point of intersection b= 0.587; standard deviation s= 0.326.
^Main phase: r = 0.889; k = 0.681 ; b = -1 .471 ; s = 1 .000.
^Final phase: r = 0.426; /c = 0. 1 1 7; b = -0. 1 95; s = 0.353.

103

Cumulative growth (%) /

^ - -.--^

— ^ V/-

/ 0 /
/ cr y /

/-/" /•

/ ''''

tj~r"

r 1 1 1

1 1 L 1 1 1 J 1 1 1 J 1 1 1

1 10 20 30 10 20 31 10 20 30 10 20 31 10 20 31 10 20 30
April Sept.

Figure 7 — Characteristic growth curve ampli-
tudes for cembran pine, spruce, and European
larch: afforestation "Haggen" in 1 ,800 m a.s.l.

Apart from the differences in the beginning and end of
the period of extension growth, cembran pine is the most
sensitive to temperature of the three species, as the high
amplitude shows. It can be concluded that the cembran
pine makes the most efficient use of warm weather condi-
tions for increment production. In the initial phase of

extension growth, cembran pine is already well ahead of
the other species in terms of growth. In the main growth
phase, larch is also surpassed by the cembran pine, while
spruce comes very close to it.

For the main growth period, the two extremes of the
growth curves for spruce and larch are almost parallel.
This shows that increment production per pentade is
more balanced between warm and cold conditions than
in the case of cembran pine, which is much more sensitive
to temperature fluctuations.

REFERENCES

Burger, H. 1926. Untersuchungen iiber das Hohen-
wachstum verschiedener Holzarten. Mitt. Schweiz.
Centralanst. Forstl. Vers, wesen; XIV/I.

Kronfuss, H. 1985. Die Zuwachsleistimg einer Hochlagen-
aufforstung mit Zirbe (Pinus cembra) auf einem
Siidhang, in Abhaangigkeit von der Seehohe. lUFRO
Meeting Proceedings. Davos, Switzerland.

Larcher, W. 1980. Okologie der Pflanzen auf
physiologischer Grundlage. Ulmer, Stuttgart.

Week, J. 1955. Forstliche Zuwachs und Ertragskimde.
Neimiann, Berlin.

104

SURVIVAL AND GROWTH OF PLANTED
CEMBRAN PINES AT THE ALPINE
TIMBERLINE

Josef Senn

Walter Schonenberger
Ueli Wasem

Abstract — In an experimental afforestation with cembran pines
{Pinus cembra L.) at the subalpine timberhne, tree survival was
mainly determined by the date of disappearance of the snow
cover in spring. Trees survived well on early snow-free sites and
poorly on sites with long-lasting snow cover. The major causes of
mortality were the two parasitic fungi Ascocalyx abietina and
Phacidium infestans. Tree height, as an index for growth, was
dependent on the amount of energy available during the vegeta-
tion season, and therefore negatively correlated with both alti-
tude above sea level and wind velocity, and positively with global
radiation.

In the severe winter of 1951-52 avalanches caused
major destruction all over the Alps. This initiated a
multidisciplinary research program to restore the upper
treeline. The aim of the program was to develop methods
to alforest treeless avalanche catchments and abandoned
meadows within the subalpine forest belt in order to es-
tablish new protection forests.

Protection forests are an important feature in alpine
landscapes. Many areas can only be permanently inhab-
ited as long as these forests remain intact. In some of
these forests, however, severe problems concerning stand
stability have become evident. Inadequate plantation
techniques in the past have frequently led to forests that
were highly susceptible to natural hazards such as
storms. On the other hand, locally dense ungulate popu-
lations prevented successful natiu-al rejuvenation in the
existing forests.

Both natural regeneration and artificial afforestation
generally have become more difficult toward the alpine
timberline, since all biological processes become slower
with increasing altitude above sea level. In general, com-
pared to natiiral invasion of forest trees, planting reduces
the time needed to establish forests that fulfill their pro-
tective function. Therefore artificial afforestation may be
highly desirable. Poor financial resources, however, may
Umit planting to the most promising sites.

The experimental afforestation at StiUberg was designed
to obtain information about site conditions that determine
success or failure of future afforestations, and to be able

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Josef Senn, Walter Schonenberger, and Ueli Wasem are members of the
High Altitude Afforestation Group at the Swiss Federal Institute for For-
est, Snow and Landscape Research, CH-8903 Birmensdorf, Switzerland.

to inform forestry personnel which sites would be most
promising for successful planting and on which sites
planting would be fruitless.

Cembran pine {Pinus cembra L.) is one of the three tim-
berline species tested at StiUberg. This species, extending
to the highest altitudes of any trees in the Alps, played an
important role in the culture of local human populations
(Holtmeier, these proceedings), produces valuable timber,
and is esteemed by "green" tourism as an important fea-
ture of natural landscapes. Nonsustainable exploitation
in medieval times and pasturalism reduced cembran pine
to a fraction of its original abundance, although there is
evidence for natural range expansion in the past 100 years.
Artificial reproduction and estabhshment of cembran pine,
on the other hand, pose many problems for practical for-
estry (Frehner and Schonenberger, these proceedings).

MATERIAL AND METHODS

The present study was conducted at StiUberg experi-
mental area, a highly structiu-ed northeast slope (fig. 1)
in the Dischma Valley near Davos, Switzerland (lat.
34°08' S., long. 171°41' E.). The area extends over 9.9 ha
between 2,080 and 2,230 m above sea level. The highest
parts reach beyond the altitudinal timberline. Before
plantation in 1975 various site aspects like slope, expo-
sure, global radiation (Turner 1966), wind conditions
(Nageli 1971), date of disappearance of snow cover in
spring (Rychetnik 1987), ntunber of snow-fi-ee days in
winter, avalanche fi'equencies, soil types (Blaser 1980),
and vegetation (mainly dwarf shrub communities) (Kuoch
1970) were recorded on a very fine scale. Schonenberger
(1975) described the site in detail. Results from the
StiUberg experimental afforestation were published by
Schonenberger and others (1988). Practical aspects for
high-altitude forestry were included in Schonenberger
and others (1990).

Three coniferous tree species were planted in 1975:
5-year-old cembran pines, 3-year-old mountain pines (P.
uncinata (Miller) Domin = Pinus montana Miller = Pinus
mugo Turra ssp. uncinata Domin) the erect form of the
mugo pine, and 1-year-old European Isirches (Larix de-
cidua Miller). Before planting, the whole area was di-
vided into square units of 3.5 by 3.5 m (fig. 1). In each
square unit 25 trees of one species were planted. The
number of trees per imit may have been reduced by ad-
verse local soil conditions like barren rock. The squares
contained alternatively cembran pines, mountain pines,
and larches. Altogether some 90,000 trees were planted
in 4,052 square units.

105

Figure 1 — Topographic model of the Stillberg
experimental area with the grid of the affores-
tation units (3.5 by 3.5 m).

This paper reports the results of survival of the cembran
pines in comparison with mountain pine and European
larch in the whole study area, and of height variation and
mortality causes in a subsample. Survival of each tree
was recorded annually as from 1976 (number of square
units with cembran pines = 1,351, with mountain pines =
1,350, and with larches = 1,351). Mean tree height, gen-
eral constitution, type and degree of abiotic and biotic
damage, and causes of mortality were determined annu-
ally in 228 square units containing cembran pines, in 226
square imits containing mountain pines, and in 226
square units containing larches. Twenty-eight qualitative
"types" of damage and potential causes of death could be
attributed in the field.

Correlation analysis was used to test for interactions
between various site parameters and tree performance.
The relative importance of site parameters was evaluated
by stepwise regression analyses (SAS 1985) for survival
and tree height. The survival rates were arcsine trans-
formed before performing stepwise regression analysis.

RESULTS
Survival

During the first 3 years after planting, survival was
highest in cembran pine compared to mountain pine and
larch, but in 1991 only 25.3 percent of the planted cembran
pines survived, compared to 37.8 percent in mountain
pine and 73.1 percent in larch (fig. 2).

Stepwise regression of survival rate in a square unit
(number of trees alive in 1991/number of trees planted in
1975) of cembran pine revealed that the date of disappear-
ance of snow cover in spring and the altitude above sea
level were the most important factors, explaining some
46 percent and 8 percent of total variation, respectively.
In our study area the survival rate was high on the early
snow-free and lower sites. Survival decreased with in-
creasing altitude, and almost no trees survived on sites
where snow lasted on 10-year average longer than June 10.

Avalanche frequency, slope inclination, global radiation,
and the number of snow-free days in winter also had sig-
nificant effects on survival, but each of these factors con-
tributed less than 1 percent to the total variation.

Mortality Causes

Of the 28 recorded causes for mortality, Ascocalyx
able- tina (Lagerb.) (= Gremmeniella abietina [Lagerb.]
Morelet), a parasitic fungus, was the major factor, killing
some 52.3 percent of the planted cembran pines. A second
fungus, Phacidium infestans Karst., the snow blight,
killed another 11.8 percent of the trees. Some 4.8 percent
of the trees died or disappeared without any detectable
reason. Undetermined fungi killed 1.3 percent and com-
petition with herbaceous vegetation some 0.6 percent of
the planted trees. Herbivores caused no losses.

We found a significantly positive correlation between the
proportion of cembran pines killed by Ascocalyx abietina
and the date of disappearance of snow cover in spring
(fig. 3). Square units that were snow free before May 12
suffered only slightly (13.9 percent of the trees killed),
whereas AscocaZyx killed 90.3 percent of the trees in units
that were covered by snow later than June 10. Occur-
rence oi Ascocalyx was also positively correlated with
altitude {r = 0.55, P < 0.0001, n = 228), the fungus kilHng
an increasing proportion of yoimg trees with increasing
altitude.

Between the proportion of cembran pines killed by Pha-
cidium infestans and the date of disappearance of snow
cover in spring, however, we found a negative correlation
(r = -0.42, P < 0.001, n = 228). This means that the high-
est proportion of trees was killed in squares that were

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91

Year

Figure 2— Survivorship curves of the three
tree species, cembran pine (Pinus cembra),
mountain pine (P. uncinata), and European
larch {Larix decidua), planted in 1975 in the
Stillberg experimental area.

106

1.0

E o

O 0)

a
o

0.8 •

0.6

0.4

0.2

• . . • • •

• • • I •

•i

•t .

• • •

t'..' «.

120 130 140 150 160 170

Date of Disappearance of Snow Cover (day)

Figure 3 — Relationship between the propor-
tion of cembran pines killed in a square unit by
Ascocalyx abietina (= number of trees killed by
Ascocalyx since 1 976/number of trees planted
in 1975) and the date of disappearance of snow
cover in spring. Day 1 30 = May 1 0; day 1 60 =
June 9. Regression line: y = -3.0637 + 0.0251 x,
r=0.8^,P« 0.0001, n = 228.

snow free early, whereas the importance of Phacidium
as a mortaHty cause decreased with increasing duration
of snow cover. Similarly the ratio of trees killed by Pha-
cidium decreased with increasing altitude {r = —0.26,
P<0.0001, n = 228).

The occurrences of the two parasitic fungi were signifi-
cantly negatively correlated within square units (r = -0.59,
P < 0.0001, n = 228). If one of the two fungus species was
common in one square, the other was rare.

DISCUSSION
Survival

Duration of snow cover in spring was found to be the
main limiting factor for survival of young cembran pines
in the upper subalpine forest belt. On sites where snow
cover lasted until after Jime 10 cembran pines were al-
most completely excluded. The second evergreen Pinus
species, the mountain pine, was similarly vulnerable to
long-lasting snow cover. The European larch, a deciduous
conifer, was much more tolerant. In this species 30 per-
cent of the trees siu^ved on sites snow covered until after
June 10. Further, survival of cembran pine was signifi-
cantly affected by altitude, although this factor was clearly
less important than duration of snow cover in spring. In-
terestingly, the impact of site factors such as global radia-
tion, number of snow-free days, slope inclination, and ava-
lanche frequency on tree survival was negligible compared
to the first two factors. Avalanche frequency, however,
will certainly become more important in the near future.
Young trees with flexible stems generally survive ava-
lanche impacts without significant damage. But with in-
creasing height and stem diameter the trees are becoming
increasingly vulnerable to stem breakage caused by snow
movements, a tendency that has become evident only dur-
ing the last years.

Mortality Causes

In our study area the two parasitic fungi Ascocalyx
abietina and Phacidium infestans were the major mortal-
ity factors for cembran pines, and these fungi still kill sig-
nificant numbers of young trees every year.

Tree Height

Mean tree height in 1991 was lowest in cembran pine
(58.8 cm) compared to moxmtain pine (72.7 cm) and larch
(59.8 cm), although initial mean height was largest in
cembran pine (fig. 4).

Variation in height was considerable, depending on
local site conditions. On sunny southeast exposed sites,
average height of cembran pines was above 80 cm, whereas
on shady sites mean height was around 45 cm. Stepwise
regression revealed that tree height was mainly affected
by altitude, global radiation, and wind velocity, these fac-
tors contributing 21 percent, 15 percent, and 10 percent
to the total variation. Duration of snow cover in spring
explained 3 percent of the variation in height. In our
study area the cembran pines were high on sites at low
altitudes, receiving high amounts of radiation and being
protected from strong winds. Slope inclination, avalanche
frequency, and number of snow-free days in winter had no
significant impact on tree height.

(0
0)

-♦■ -

- C€

- M<

- La

>mb
}unt
rch

'an pine
ain pine

—

-«-••

✓

r''

b-'

✓

...I

p: —

— yt
('

1 1

—

r--

..-1

r''

1 .-

■

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91

Year

Figure 4 — Annual mean height of the three tree
species, cembran pine {Pinus cembra), mountain
pine (P. uncinata), and European larch (Larix de-
cidua), planted in 1975 in the Stillberg experimen-
tal area.

107

Ascocalyx infestations first become evident in spring
in the dying buds. Later in the season the shoots die back
along the branches. The fungus may eventually kill a
young cembran pine after 2 or 3 years. The close relation-
ship between the date of disappearance of snow cover in
spring and the occurrence Ascocalyx (fig. 3) indicates
that the fungus preferentially attacks weakened plants.
On the sites with long-lasting snow cover, the vegetation
season is relatively short, and thick layers of raw himius
have accumulated because of the low temperatures.
Therefore, they are poor sites for tree growth. Similarly,
Kurkela (1984) found that the trees most heavily affected
hy Ascocalyx had the lowest growth rate before the epi-
demic. He explained the differences in tree growth by the
different site fertility. In our study the hypothesis that
Ascocalyx preferentially attacked weakened trees was fur-
ther supported by the finding that the proportion of trees
killed by the fungus significantly increased with increas-
ing altitude (toward the altitudinal tree line). In our study
area Ascocalyx attacked mountain pines, too, where it also
caused considerable losses.

In contrast to the situation in Ascocalyx, the proportion
of cembran pines killed by Phacidium was negatively cor-
related with the date of disappearance of snow cover and
with the altitude. According to this correlation the trees
growing at lower altitudes and on sites that were snow
free early were killed more frequently by Phacidium than
the trees growing at higher altitudes and on sites with
long-lasting snow cover. Young trees that looked healthy
in one year were often found dead in the following spring.
Interestingly, Roll-Hansen and others (1992) reported
from Scandinavia that the most vigorously growing plants
were most susceptible to Phacidium infestations, whereas
poorly developing plants were most resistant.

The fact that Phacidium killed the highest proportion
of trees at early snow-free sites seems contradictory to
the biology of the fungus, that exclusively attacks parts
of trees that are covered by snow (Roll-Hansen 1989).
Branches destroyed by this fungus are clearly visible after
disappearance of snow cover and mark the prevalent snow
depth. Our study area contains some older cembran pines
growing on ridges (at the sites that are snow free early).
These trees originating from caches of the nutcracker
{Nucifraga caryocatactes) host Phacidium, which is non-
lethal for the large trees that extend considerably beyond
the snow cover. These trees may act as sources for local
Phacidium outbreaks that eventually kill the young trees
completely covered by snow. Since this fungus specifically
attacks cembran pine and does not occur on mountain
pine (but see Roll-Hansen 1989), it may not spread as eas-
ily over larger areas as does Ascocalyx, which uses moun-
tain pine as an alternative host. Therefore, even heavier
outbreaks of Phacidium remain locally restricted around
mature cembran pines. Together with the fact that Pha-
cidium preferentially attacks the most vigorous trees, this
may explain the seemingly paradoxical negative correla-
tion between duration of snow cover in spring and occur-
rence o{ Phacidium.

On sites with high avalanche frequencies the young
trees were partially excluded, not by stem breakage or
uprooting, but by competitive interactions with herba-
ceous vegetation. The moving snow carried rich mineral

soil leading to a lush herbaceous vegetation that out-
competed the slowly growing young cembran pines. Ava-
lanches, however, become a more serious problem for the
surviving trees as they increase in height and stem diam-
eter. Broken stems were found more frequently in recent
years.

Herbivore impact on the survival of young cembran pines
was relatively unimportant. Herbivores killed no trees, al-
though they may reduce tree growth. Browsing vertebrates
like black grouse (Tetrao tetrix) and chamois (Rupicapra
rupicapra), however, clearly preferred mountain pine and
larch to cembran pine. Black grouse caused some locally
restricted damage by browsing buds and first-year needles.
They obviously preferred to browse on easily accessible
trees growing on ridges with little snow cover (Streule
1973). Populations of herbivorous insects like phloem-
sucking aphids (Cinara cembrae Seitner, see Grbic, these
proceedings, and Pineus cembrae [Choi.] Amand.) fluctu-
ated in size but were low in most of the years when com-
pared to the other two tree species. Although visible dam-
age by herbivores was relatively rare, much rarer than
fungal infections, the herbivores may have indirectly af-
fected the survival of the trees by injuring plant tissue,
making the trees more susceptible to infestations by fungi.

Tree Height

Tree height in cembran pine was clearly dependent on
the available amount of energy during the vegetation sea-
son. Altitude above sea level, global radiation, and wind
velocity explained the largest amount of variation in
height. We are aware that tree height does not directly
represent tree growth, since we did not measure annual
growth. We only recorded tree height in summer, exclud-
ing the growing shoot of the current year. Tree height,
which is the result of annual growth minus the losses
through damage in the apical region, may increase or
decrease from one year to another. Sixteen years after
plantation, differences in average tree height among sites,
however, are primarily the result of variation in growth
caused by variation in local site conditions.

The date of disappearance of snow cover in spring, al-
though significantly related to tree height, explained only
3 percent of the variation in height in contrast to survival
where the snow conditions in spring mainly explained the
observed pattern.

Slope inclination and avalanche frequency, which may
be related to stem breakage, had no detectable impact on
tree height. Further, we found no relationship between
the number of snow-free days in winter and tree height,
although trees that were snow free during extended peri-
ods in winter should have been more vulnerable to herbi-
vore browsing than trees that were covered by deep snow.

CONCLUSIONS

Foresters should carefully select favorable microsites
when planting cembran pines in high-altitude afforestations.
Particular attention should be paid to the spatial pattern
of snowmelt in spring, and no cembran pines should be
planted at sites with prolonged duration of snow cover.

108

REFERENCES

Blaser, P. 1980. Der Boden als Standortsfaktor bei
Aufforstiingen in der subalpinen Stufe (Stillberg,
Davos). Eidgenossische Anstalt fur das forstliche
Versuchswesen, Mitteilimgen. 56(3): 527-611.

Kuoch, R. 1970. Die Vegetation auf Stillberg (Dischmatal,
Kt. Graubiinden). Eidgenossische Anstalt fiir das
forstliche Versuchswesen, Mitteilungen. 46(4): 329-342.

Kurkela, T. 1984. The growth of trees affected by Grem-
meniella abietina. In: Manion, P. D., ed. Scleroderris
canker of conifers. The Hague: Martinus NijhofiEDr W.
Junk Pubhshers: 177-180.

Nageh, W. 1971. Der Wind als Standortsfaktor bei
Aufforstungen in der subalpinen Stufe (Stillbergalp
im Dischmatal, Kanton Graubiinden). Eidgenossische
Anstalt fiir das forstliche Versuchswesen, Mitteilungen.
47(2): 33-147.

Roll-Hansen, F. 1989. Phacidium infestans: a literature
review. European Journal of Forest Pathology. 19:
237-250.

Roll-Hansen, F.; Roll-Hansen, H.; Skroppa, T. 1992.
Gremmeniella abietina, Phacidium infestans, and other
causes of damage in alpine, young pine plantations in
Norway. European Journal of Forest Pathology. 22:
77-94.

Rychetnik, J. 1987. Snow cover disappearance as influ-
enced by site conditions, snow distribution, and ava-
lanche activity. In: Proceedings of the international
symposiima on avalanche formation, movement and
effects; Davos 1986. lAHS-Pubhcation Nr. 162.

SAS. 1985. SAS user's guide: statistics. Gary, NC: SAS
Institute.

Schonenberger, W. 1975. Standorteinfliisse auf
Versuchsaufforstimgen an der alpinen Waldgrenze
(Stillberg, Davos). Eidgenossische Anstalt fiir das
forstliche Versuchswesen, Mitteilungen. 51(4): 357-428.

Schonenberger, W. 1985. Performance of a high altitude
afforestation under various site conditions. In: T\irner,
H.; Tranquillini, W., eds. Establishment and tending of
subalpine forest: research and management: Proceed-
ings of the 3rd lUFRO workshop P 1.07-00, 1984.
Eidgenossische Anstalt fiir das forstUche Versuchswesen,
Berichte. 270: 233-240.

Schonenberger, W.; Frey, W., eds. 1988. Untersuchimgen
zur Okologie und Technik der Hochlagenaufforstung;
Forschungsergebnisse aus dem Lawinenanrissgebiet
Stillberg. Schweizerische Zeitschrift fiir Forstwesen.
139: 735-820.

Schonenberger, W.; Frey, W.; Leuenberger, F. 1990.
Okologie und Technik der Aufforstung im Gebirge -
Anregimgen fiir die Praxis. Eidgenossische Anstalt fiir
das forstliche Versuchswesen, Berichte Nr. 325. 58 p.
[Version franfaise: Ecologie et technique des afforesta-
tions en montagne - Suggestions a I'usage des praticiens.
Traduction du rapport No. 325.] [Versione italiana:
Ecologia del rimboschimento di montagna e pratica
della tecnica di piantagione. Traduzione del rapporto
No. 325.]

Streule, A. 1973. Schaden in Gebirgsaufforstungen durch
das Birkhuhn (Lyrurus tetrix). Biindnerwald. 28(8):
249-254.

Tiimer, H. 1966. Die globale Hangbestrahlung als
Standortsfaktor in der subalpine St\ife. Eidgenossische
Anstalt fiir das forstliche Versuchswesen, Mitteilungen.
42(3): 111-168.

109

Influence of
Environmental Factors

International Workshop
St. Moritz 1 992

s ^ '

110

CHANGES OF SWISS STONE PINE
APHID LIFE CYCLE, DENSITY, AND
POPULATION STRUCTURE IN HIGH-
ALTITUDE SWISS STONE PINE
AFFORESTATION

Mihailo Grbic

Abstract — The order of form appearance, structure, and number
of individuals in colonies of Cinara cembrae on Swiss stone pine
{Pinus cembra) used for afforestation in the Dischma Valley were
strongly related to the slope aspect. The abnormality was par-
ticularly apparent on north slopes. North slope colonies
showed incomplete life cycles. All insects seen there were
virginoparae and their larvae. Colonies were small in number
and were developed from winged virginoparae from other areas.

The Alps are of great importance to Switzerland as an
area for hviman habitation and as a health and holiday re-
sort. Afforestation in high altitudes is costly and involves
high risk, as many failures occur. Because of the slow de-
velopment of such afforestation it is difficult to under-
stand and to determine the causes and processes that lead
to a failure and then to suggest measures to improve
success.

Environmental factors of the upper timberline area are
totally different from those of the forest zones. These
harsh conditions affect the specific growth and develop-
ment of planted species; for example, low survival rate
and low annual increment. The surviving plants, depend-
ing on the biological characteristics of the species, are
more or less open to insect and fungal attack. These same
high-altitude factors (for example, low temperature, short
growing period, and short day photoperiod) tend to change
the life cycle, density, and population structure of insects.
This is particularly true for aphids due to their polymor-
phic nature, and their ability to alter the nimaber of gen-
erations per season.

STUDY AREAS AND METHODS

These investigations were carried out in field plots of
the Stillberg research area and the Lucksalp comparative
afforestation area, in the timberline area of the Dischma
Valley (Canton Graubunden). The period of the study
was between late May and the middle of September,
which is the time of greatest insect activity.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Mihailo Grbic is Teacher at the University of Belgrade, Faculty of For-
estry, Kneza Viseslava 1, YU-11030 Belgrade, Yugoslavia.

Stillberg Hes at 2,080 to 2,230 m above sea level (a.s.l.),
with a northeast aspect, and 30° to 45° slope. The slope is
divided by spurs, and as a result there are three different
aspects (north-, northeast-, and east-facing slopes). On
the opposite side of Dischma Valley lies Lucksalp at
2,200 m a.s.l., with a southwest aspect.

The subjects of investigation were 19-year-old Swiss
stone pine (Pinus cembra L.) and the Swiss stone pine
aphid (Cinara cembrae Seithner), the most frequently ob-
served insect pest on that afforestation species. The trees
were planted 70 cm apart in a grid pattern in approxi-
mately 4,000 square plots (3.5 by 3.5 m). The Swiss stone
pine stocks were alternated from plot to plot with two
other afforestation species, 17-year-old Swiss mountain
pine (Pinus mugo ssp. uncinata [Mill.] Domin.), and
15-year-old European larch (Larix decidua Mill.), each
plot containing 25 trees (Shonenberger 1985).

The samples of Swiss stone pine aphid were identified
by comparing them to recent accounts from various parts
of Europe (Carter and Maslen 1982; Eastop 1972; Pintera
1966; Stroyan 1955, 1960). The structure of antennae,
rostrum, abdomen, and hind legs of collected material
were compared with these references.

The life cycle was observed in colonies feeding on par-
ticular sample trees. On Lucksalp, 30 trees (5 trees x
6 plots) were regtdarly examined, and 50 (5 trees x
10 plots) were examined on Stillberg. Observations were
made to establish the relationship between the intensity
of infestation and site aspects.

The trees were examined every 10 days to determine
density of pest population and structvire of colonies.

MORPHOLOGY

The morphological characteristics of specimens found
on Stillberg and Lucksalp (tables 1 and 2) are similar to
the central European description (Pintera 1966). The
chronological series of morphs is: fundatrix, aptera
virginopara, alata virginopara, ovipara, and aptera and
alata males.

Compared with the apterae of the succeeding genera-
tions, fundatrix shows "fundatrix facies" characteristics
(Lees 1961). With the Swiss stone pine aphid, the anten-
nae and legs are relatively shorter in comparison to the
body. The processus terminalis is shorter with fewer
rhinaria. Features similar to the fundatrix facies that
were observed by Hille Ris Lambers (1955), Lees (1961),
and Stroyan (1960) on nonfimdatrix morphs of some

111

Table 1 — Biometric data for different morphs of Cinara cembrae from Lucksalp

Biometric data

Aptarae
virginoparae

Alatea
virginoparae

Morph
Apterae
male

Alatae
male

Sexual
female

Body length (mm)

Length of antennal segments
II
IV
V

VI (base + proces. termlnalls)

Length of rostal segments (^)
IV
V

Hind tarsus segments (^i)
I

basal diameter
dorsal length
ventral length

length

Hind tibia length (mm)

Length of longest hair on
III ant. segment
hind tibia

VIII abdom. tergite

Number of hairs on
II ant. segment
VI ant. segment
base

proces. term,
ultimate rostral segment

Number of hairs on
subgenital plate
VIII abdom. tergite

Number of secondary rhinaria on
ant. segments

III

3.05-3.52

680-720
300-340
400-410
167-180 +
75-83

218-238
100-112

74- 78
95-105

155-182

385-420
3.20-3.30

75- 97
75-120

110-165

9-15

9- 13
4-5
6-8

10- 12
4-6

0-2
0-2
2-3

3.00-3.10

635-650
275-290
320-345
150-160 +
59-66

170-182
85-88

57-62
62-66
122-130

330-350
2.90-3.05

88-92
70-105
185-200

6-7

10-12
3-4
8

9-11
2-3
5

2.50-3.30

830-880
335-400
430-500
190-198 +
75-90

230-250
90-110

60-70
98-110
150-180

340-390
1.85-3.00

75-120
75-110
100-112

10-15

10-12
5

10-12

10-12
7-11

10-13
7-10
4-5

2.40-3.00

845-910
360-430
440-500
109-198 +
85-95

220-230
105-110

60-65
85-105
150-180

385-400
2.90-3.40

115-140
140-180
90-120

7- 14

8- 10

5- 6
8-10

6- 9
6-9

55-83
10-13
5

3.80-4.70

830-890
360-385
480-500
192-212 +
75-80

245-250
106-118

68-88
105-120
187-200

370-420
3.70-3.78

95-123
112-145
120-138

10-16

8- 13
5-6

9- 11

9-13
7-8

2-3
1-2
4-7

aphids that live at low temperatures were not recorded in
these study areas.

Morphological differences between apterous and alatae
virginoparae are found not only in the presence or ab-
sence of the wings, but also in the following characteris-
tics. In alatae:

• Segmentation between head and prothorEix is more
conspicuous.

• Compound eyes are bigger, v(ath many facets.

• Antennae bear five times as many rhinaria.

• The body is shorter, and the mesothoracic terga con-
sists of a compound notum and postnotum.

• The scutum occupies the central major part of the
nottun and is divided in two mesothoracic lobes, which are
hardly developed.

The virginoparae of every generation are morphologi-
cally uniform, and under field conditions it is impossible
to find out how many generations have developed. Also,
the length of time required for passing through four in-
stars, from birth to adult, is variable and dependent on
two external factors (food quality and temperature) and
two internal factors (birth weight and whether the morph
is winged or unwdnged) (Dixon 1987). This is the reason
for the presence of different generations of the same form
occurring simultaneously in one colony.

As is known, Swiss stone pine aphids produce both
alatae and apterous males. Both forms were found in the
research areas, but with a preponderance of the wingless
form. Both forms of males are readily recognized by their
sclerotized genital structure. The body, especially abdo-
men, is smaller and more slender, and the antennae bear
more rhinaria.

112

Table 2 — Biometric data for different morphs of Cinara cembrae from Stillberg

Morph

Aptarae

Alatea

Apterae

Alatae

Sexual

Biometric data

virginoparae

male

female

Body length (mm)

2.90-3.70

2.90-3.90

2.50-3.10

2.40-3.00

3.90-4.70

Length of antennal segments

II
II

540-800

^ Art rt C rt

630-650

820-800

840-930

800-860

260-355

250-280

320-400

360-440

350-400

340-450

www OOU

tOU-wuU

ARn-t^ 1 n

HOU 0 1 u

u-ouu

VI (base + proces. termlnalis)

170-188-1-

153-167 +

183-210 +

192-200 +

175-222 +

80-85

58-68

75-95

88-95

75-85

Length of rostal segments (n)

rtOQ OCA

238-250

1 62-1 72

235-250

220-225

235-250

100-108

83-87

95-102

100-105

105-118

Hind tarsus segments (n)
1

h^^^sl dif^mptAr

75-78

WW w9

DU-DO

DO- 1 UO

dorsal length

88-108

60-62

95-1 02

88-1 07

105-125

ventral length

158-175

122-127

150-178

150-170

185-230

length

365-385

300-340

335-395

380-400

420-450

Hind tibia length (mm)

2.40-3.39

2.82-2.98

1 .78-2.98

2.95-3.37

3.70-3.88

Length of longest hair on

III ant. segment

60-00

85-88

78-125

1 1 2-1 42

95-130

hind tibia

60-105

75-1 10

78-1 12

140-175

1 12-150

VIII aooom. tergite

168-220

195-210

100-115

88-130

112-138

Number of hairs on

II ant. segment

11-15

6-7

10-16

7-15

10-18

VI ant. segment

base

8-12

9-13

10-13

8-1 1

8-16

proces. term.

4-5

3-4

5-6

4-6

ultimate rostral segment

8-9

7-9

9-12

8-11

Number of hairs on

subgenital plate

9-11

9-12

6-9

9-14

VIII abdom. tergite

4-6

7-12

6-9

Number of secondary rhinaria on

ant. segments

III

0-4

8-10

10-15

50-89

2-4

2-3

1-3

7-11

11-15

0-2

3-4

4-6

5-6

2-7

Alatae and apterous males and alatae virginoparae pos-
sess a far larger number of secondary rhinaria than apter-
ous virginoparae and sexuales females. The following
numbers of secondary rhinaria on the third antennal seg-
ment were found: apterous virginoparae 0-4, alatae
virginoparae 8-11, apterous males 10-15, and alatae
males 50-89. The number of secondary rhinaria on
sexuales females was similar to that found in apterous
virginoparae.

The function of secondary rhinaria is probably different
in male and female individuals. The secondary rhinaria
in males appear to be the main receptors of sex phero-
mones, while secondary rhinaria of alatae virginoparae
must have a function other than pheromone detection.
This function is possibly host selection, as has been shown
by many authors with other aphid species (Alikhan 1960;
Chapman and others 1981; and Pospisil 1976).

Sexuales females resemble the apterous virginoparae,
but may be somewhat bigger. The tibiae of the hind legs
are often longer with longer setae also, and numerovis
pseudosensoria on them.

LIFE CYCLE

General — The life cycle of the Swiss stone pine aphid is
strongly related to site aspects and environmental factors.
Generally, the Swiss stone pine aphid has a monoecious
holocyclic type of life cycle. The life cycle starts in May
v^dth the hatching of fundatrices larvae from eggs laid in
rows on the needles. In the covirse of svmimer a small
niunber of generations of parthenogenetic females
develop.

Generations of winged parthenogenetic viviparous fe-
males arise from colonies of apterae later in the summer.

113

Sexuales forms appear in the second half of August. This
early development of sexuales is due to the alpine climate.
In the low-lying positions where Swiss stone pine has
been artificially introduced, Swiss stone pine aphid has
never been found.

Observed — The life cycle observations in the Dischma
Valley show the following characteristics: colonies from
Lucksalp and all slopes of Stillberg, except north-facing
slopes, show complete life cycles with six morphs. Par-
allel appearance of two forms of males, observed by
Seithner (1936), is a unique case in the family Lachnidae.
The north-slope colonies showed incomplete life cycles.
All insects seen there were virginoparae apterae or alatae
and their larvae. Colonies were small in number and
were developed from winged virginoparae from other
areas that colonized plants on north-facing slopes from
other areas (table 3).

The order of appearance of particular forms on
Lucksalp was earlier than on the Stillberg research area.
The first observation, made in the last third of May,
shows 68 percent unhatched eggs, and all fundatrix lar-
vae were part of the first two larval instars. At the same
time on Stillberg 87 to 90 percent of the eggs were
unhatched.

In the first third of June the colony structure on
Lucksalp was: 7 percent of the eggs were unhatched,
young fundatrices in all four larval instars were found,
but no adults. A similar, but slightly different, situation
was recorded on Stillberg. There were 25 to 38 percent
unhatched eggs, and the colonies contained only the first
three larval instars.

During the observations in the second third of June we
found adult fundatrices on Lucksalp and on the east
slopes of Stillberg. With the appearance of fundatrices,
the density of populations increased gradually.

On the other Stillberg slopes (ENE and NE) fundatrices
were found 10 days later. Simultaneously we recorded
adults of apterous virginoparae on Lucksalp. This is a
time of heavy increases in population, because the multi-
plication rate of apterous virginoparae is very high, and
the duration of larval instars very short.

In an experiment that was conducted under conditions
of changeable room temperature, one apterous virgino-
parae was isolated on a twig that was kept moist. Ten
days later, we recorded six apterous virginoparae of
a new generation (which started with larviposition); 16
in the fourth larval instar; and 18 larvae were in yoimger
stadiums.

It was presumed that the effect of the multiplication
rate is not the same in nonisolated field conditions, but
on some trees the increments of population for 10 days
increased by a factor of 10 or more. On the other hand,
some other populations were substantially reduced by the
action of parasites and predators.

In the first third of July at Lucksalp a typical popula-
tion consisted of more than 170 individuals; almost all
belonged to apterous virginoparae larvae (83 percent) or
adult insects ( 15 percent). The rest (2 percent) were
alatoid larvae of alatae virginoparae. The fundatrix gen-
eration was dead. At Stillberg, populations were smaller

with most of the population made up of larvae and adults
of apterous virginoparae; however, fundatrices were also
found to be present. The fourth larval instar of alatae
virginoparae was recorded on the east slopes only. The
proportion of alatoid larvae was 10 percent of the total
number of the fourth larval instar.

The first Swiss stone pine aphid was recorded on the
north slopes of Stillberg in the middle of July. Colonies
were small and consisted of only a few adult alatae
virginoparae and young larvae. Few of the colonies were
without adults. The population structures of the other
slopes were unchanged in content until the last third of
August. Only the numbers of individuals were variable,
depending on weather conditions.

The rapid decrease in number of individuals at the be-
ginning of August was caused by low temperatures and
snow. The greatest changes in population quantity were
recorded on north slopes. The average population struc-
ture in the course of July and August was: The majority
of individuals belonged to various larval instars, and the
adult part of the population consisted of apterae and
alatae viviparous parthenogenetic females. Average ratio
of winged and wingless forms was 1:5.3 with an increas-
ing tendency to winged forms in colder sites. This ten-
dency is due to the strategy of the species to avoid expo-
sure to more extreme climatic conditions (by colonizing
another host plant of the same species, but in better envi-
ronmental conditions). Extremes of the ratio were 1:8 and
1:2. The first was recorded on Lucksalp, and the second
on Stillberg north slopes.

The highest population numbers were found in the last
third of July, after which the numbers decreased rapidly
as a consequence of low temperatures at the beginning
of August. Thereafter, the population number increased
quickly again to a secondary peak at the end of August.
Occurrence of the last generation was recorded after an-
other cool weather period at the end of August. As the
overwintering eggs are produced sexually, the last genera-
tion of the year must contain males and sexual females,
instead of the parthenogenetic forms that occur during
the rest of the year. However, sexuales do not occur un-
der all investigated conditions. On the north slopes of
Stillberg, for example, no sexuales forms were recorded.
On Lucksalp, occurrence of sexuales forms was about
10 days before Stillberg.

The first appearance of eggs was observed at the begin-
ning of September. Oviparous females move away from
the colony to the top part of current shoots, and lay over-
wintering eggs in rows on needle clusters. Eggs were
brown and later black and shiny 48 hours after being laid.
Average number of eggs in the row along one needle was
8.2 and the extremes were 16 and 1. Maximum egg num-
ber per one colony was 316, recorded in the first week of
September at Lucksalp. The same colony contained eight
sexuales females. Results of dissection of females show
12 ripe eggs in the ovariole on average. This was smaller
in comparison with the embryo numbers obtained by dis-
section of parthenogenetic females during July, which
was 30.

114

Table 3 — Average number of different morphs of Cinara cembrae on Swiss stone pine sample trees from last third of May to second third of
September

Morph

May

June

Julv

August

September

3/3

1/3

2/3

3/3

1/3 2/3

3/3

1/3

2/3

3/3

1/3

2/3

Lucksalp (SW slope)

Eqqs

5.6

0.6

60.0

93.0

Larvae

2.6

7.6

35.0

103.0

148.0 178.0

199.0

68.0

82.0

124.0

9.0

11.0

Fundatrices

1.2

1.5

Apterae virginoparae

2.0

27.0 12.0

24.0

8.3

18.0

24.0

7.3

6.7

Aiatae virginoparae

1.5

2.4

1.8

1.5

2.0

0.2

Oviparous females

0.1

4.9

5.4

Apterae males

0.1

0.3

1.4

Aiatae males

0.1

Stillberg (E slope)

Eqqs

7.2

2.0

0.2

3.8

57.0

Larvae

0.8

4.4

30.0

41.0

58.0 100.0

110.0

54.0

80.0

104.0

5.9

8.1

Fundatrices

0.6

1.8

1.0

Apterae virginoparae

3.6 30.0

31.0

6.3

11.0

20.0

2.9

3.2

Aiatae virginoparae

3.0

5.0

5.8

0.7

Oviparous females

0.2

2.3

Apterae males

0.1

0.3

Aiatae males

0.1

0.2

Stillberg (ENE slope)

Eqqs

7.0

3.0

1.1

Larvae

0.8

4.8

6.7

34.0

124.0 140.0

145.0

54.0

69.0

93.0

4.6

8.3

Fundatrices

2.4

0.2

Apterae virginoparae

5.0 12.0

18.0

3.2

10.0

17.0

3.0

3.5

Aiatae virginoparae

2.6

1.0

3.8

4.7

1.8

Oviparous females

1.4

5.2

Apterae males

0.8

Aiatae males

0.2

0.5

Stillberg (NE slope)

Eggs

6.5

2.0

8.0

Larvae

1.0

5.9

7.7

32.0

40.0 89.0

118.0

26.0

40.0

58.0

8.6

12.0

Fundatrices

2.3

0.8

Apterae virginoparae

1.8 10.0

18.0

2.3

7.1

9.4

3.8

9.2

Aiatae virginoparae

0.9

1.0

1.3

3.0

2.6

2.8

Oviparous females

0.8

2.0

Apterae males

0.7

1.5

Aiatae males

Stillberg (N slope)

Eggs

Larvae

5.8

43.0

1.0

11.0

18.0

Fundatrices

Apterae virginoparae

6.2

1.0

2.0

4.8

Aiatae virginoparae

1.0

3.4

1.6

1.2

Oviparous females
Apterae males
Aiatae males

115

SEASONAL MOVEMENTS

The most important factors affecting aphid feeding and
nutrition are the change, in time, of growth and develop-
ment of the host plant (Klungauf 1987). In conjmiction
with this change, seasonal movements of colonies were
observed.

The fundatrix generation moves from the needles where
overwintering eggs were laid. Feeding sites of fimda-
trices, and later generations of virginoparae, in the course
of Jime and first half of July were on last year's shoots.
During July, colonies moved to twigs and branches in
lower parts of the crown (in extreme cases some of the
colonies were fovmd on stems). This was just after the
peak of the annual increment of the shoots, when they
start to become lignified. It seems to be related to de-
scending sap movement. Conifer lachnids, as phloem
feeders, exploit food sources in lower parts of the plant
that become available as the season progresses.

The alatoid larvae of alatae virginoparae usually move
from the colony to the tops of the shoots before their final
moult. Here they moult into alatae adults and fly off to
colonize similar positions on other plants.

Simultaneously with the occurrence of sexuales, the
colonies moved ascendentally. At the end of the season
the majority of the individuals were on last year's and
current season's shoots. From this position oviparous
females moved to lay eggs on needles.

Additional trials were carried out with colonies isolated
by polytene covers, which prevented migration from top to
bottom parts of the host plant. In the beginning, the den-
sity of isolated colonies was much higher than the density
of nearby nonisolated colonies. However, early occurrence
of alatae forms indicated an attempt to change the colo-
nies' position.

Later, complete starvation of the population was caused
by restriction of movement. It was concluded that the
life cycle of Swiss stone pine aphid is very much synchro-
nized with the growth and development characteristics of
their host plant, and in conjunction with this, quality of
nutrition.

PEST DAMAGES

Keen (1938) reported that several species of lachnids
cause considerable injury to conifers in North America.
Fumiss and Carolin (1977) described visible signs of in-
jury, such as yellowing of needles or reduction of growth
of young trees. According to Seithner (1936), the damage
to Swiss stone pine is quite iinimportant, but he believed
that the trees could be weakened by the sucking and that
secondary pests (for example, Pissodes and Pityophthorus)
could easily attack them.

We found no direct visible signs of damage by Swiss
stone pine aphid except for some discolorations on feeding
sites. Our observations show that retardation of the
growth increment by Swiss stone pine aphid, if any, is
much less significant than retardation due to other envi-
ronmental factors.

Indirectly, Swiss stone pine aphid's honeydew excre-
tion, in the absence of heavy rain, accumulates on the
surface of the shoots and forms an ideal substrate for

saprophytic sooty mould fungi to develop. A black cover-
ing of hyphae and spores can reduce the quality of trees
that are grown in parks, but for the Stillberg and
Lucksalp research areas it is not so important. On the
other hand, honey derived from honeydew, particularly
from conifer lachnids, is highly valued and of considerable
economic significance. Fossel (1971) found that the Swiss
stone pine aphid produced copious honeydew in the movm-
tainous districts of Austria.

CONCLUSIONS

Information gathered during this study has allowed the
following conclusions to be drawn:

• Swiss stone pine aphid was the most common insect
pest on Swiss stone pine used for afforestation in the re-
search areas of the Dischma Valley.

• The morphological characteristics of specimens and
the chronological series of morphs found on Stillberg and
Lucksalp are similar to the conventional descriptions.
However, for some site aspects of the Dischma Valley the
life cycle of Swiss stone pine aphid was incomplete. The
life cycle abnormality was observed on the north slopes of
Stillberg.

• The order of appearance of particular insect forms,
structure, and number of individuals in colonies also were
strongly related to site aspects. This is particiilarly true
for the temperature effect that caused later development
and smaller sizes of colonies on the north slopes.

• Swiss stone pine aphid was observed to cause only
unimportant discolorations on the feeding sites. The
damage to the trees is much less significant than that pro-
duced by the normally harsh environmental conditions of
the upper timberline area.

When looked at in isolation, the damage produced by
the pest observed in this study cannot be considered seri-
ous. However, it is only one link in a chain of factors det-
rimental to the growth of afforestation plants. How the
insect interacts with other factors harmful to tree growth
(for example, fungi) is as yet unknown, and therefore the
damage caused by colonies of Swiss stone pine aphid can-
not be accurately assessed.

Until it is known how detrimental the aphid is, the im-
portance of control measures or predators and parasites
also cannot be estimated.

Clearly, 6 months investigation is too short for con-
clusive results. However, the information gained has
provided an important first step in imderstanding the
relationship between the insect pest, Swiss stone pine
aphid, and the host plant, Swiss stone pine, in high-
altitude afforestation.

REFERENCES

Alikhan, M. A. 1960. The experimental study of chemotac-
tic basis of host specificity in a phytophagous insect.
Aphis fabae Scop. (Aphididae: Homoptera). Annales
Univ. Mariae Curie - Sklodowska, Sectio C. 15: 117-158.

Carter, C. I.; Maslen, N. R. 1982. Conifer lachnids in
Britain. Alice Holt Lodge, Farnham, Surrey, Engleind:
Forestry Commission Research Station.

116

Chapman, R. F.; Bemays, E. A.; Simpson, S. J. 1981.
Attraction and repulsion of the aphid, Cavariella
aegopodii, by plant odors. Journal of Chemical Ecology.
7: 881-888.

Dixon, A. F. G. 1987. Parthenogenetic reproduction and
the rate of increase in aphids. In: Aphids: their biology,
natural enemies, and control. 2A: 269-287.

Eastop, V. F. 1972. A taxonomic review of the species
of Cinara Curtis occurring in Britain (Hemiptera:
Aphididae). Bulletin of the British Museimi, Natiiral
History (Entomology). 27: 103-186.

Fossel, A. 1971. New observation on Cinara cembrae
(Homoptera, Lachnidae). Annales Zoologici. 28: 353-365.

Fumiss, R. L.; Carolin, V. M. 1977. Western forest in-
sects. Misc. Publ. 1339. Washington, DC: U.S. Depart-
ment of Agriculture, Forest Service. 654 p.

Hille Ris Lambers, D. 1955. Hemiptera 2. Aphididae.
Zoology of Iceland. 3(52a): 1-29.

Keen, F. P. 1938. Insect enemies of western forests. Misc.
Publ. 273. Washington, DC: U.S. Department of
Agriculture.

Klingauf, F. A. 1987. Feeding, adaptation and excretion.
In: Aphids: their biology, natural enemies, and control.
2A: 225-253.

Lees, A. D. 1961. Clonal polymorphism in aphids. In: In-
sect polymorphism. Sjmiposia of the Royal Entomologi-
cal Society of London. 1: 68-79.

Pintera, A. 1966. Revision of the genus Cinara Curt.
(Aphidodea, Lachnidae) in Middle Europe. Acta ent.
bohemoslov. 63: 281-321.

Pospisil, J. 1976. Sensory, olfactory and visual orientation
of insects of the tropical region III. Visual and olfactory
orientation of alate forms of the aphid, Aphis gossypii
Glover, in Cuba. Academia de Ciencias de Cuba, Serie
Biologica. 66: 17-23.

Schonenberger, W. 1984. Performance of a high altitude
afforestation under various site conditions. Proc. 3d
lUFRO Workshop: 233-240.

Seither, M. 1936. Lachnus cembrae an. sp., die Zirben-
blattlaus. Centralbl. ges. Forstwes. 62: 33-49.

Stroyan, H. L. G. 1955. Recent additions to the British
aphid faima. Part II. Transactions of the Royal Entomo-
logical Society of London. 106: 283-340.

Stroyan, H. L. G. 1960. Three new subspecies of aphids
from Iceland (Hem., Hom.). Entomologiske Meddelelser.
29: 250-265.

117

GENETIC CONSEQUENCES AND
RESEARCH CHALLENGES OF BLISTER
RUST IN WHITEBARK PINE FORESTS

Raymond J. Ho^
Susan K. Hagle
Richard G. Krebill

Abstract — Susceptibility of whitebark pine {Pinus albicaulis) to
blister rust (caused by Cronartium ribicola) is reviewed. Prog-
ress is reported on studies that assess the level of susceptibility
over its entire range and the existence of resistance in various
stands. Two breeding approaches are discussed: (1) the tradi-
tional, where trees are selected, tested, then established in seed
orchards; (2) a natural approach that aids natural processes to
establish "a natural selection stand."

In 1910, eastern white pine (Pinus strobus) seedlings
that had been grown in Europe were planted in the area
of Point Grey, BC, Canada. Some of these seedlings were
infected with white pine blister rust (caused by the fungus
Cronartium ribicola). The disease was finally noticed in
the fall of 1921, when it was observed on endemic western
white pine (Pinus monticola). The destructive nature of
this disease in North American white pines had already
been documented in Europe (Spaulding 1911).

What followed was a frantic rush to stop blister rust's
spread by destroying infected trees, eradicating currants
and gooseberries (genus Ribes) that are alternate hosts for
the disease, and using several promising fungicides. But
this work failed, and the fungus kept spreading. By about
1960, the fungus had spread throughout most the range
of whitebark pine (Pinus albicaulis) (fig. 1), which also in-
cludes most of the range of western white pine and sugar
pine (Pinus lambertiana). Examples of blister rust on
whitebark pine are illustrated in figure 2.

The first infected whitebark pine recorded was in
the arboretum of the University of British Columbia,
Vancouver, BC, in 1922 (Bedwell and Childs 1943). The
first discovery of blister rust on native whitebark pine
was in 1926 on the Birkenhead River in the Coast Range
of British Columbia, Canada (Childs and others 1938).
Intense infection and mortality by blister rust was not far
behind (Bedwell and Childs 1943).

The degree of infection was 100 percent in many stands
in the northern areas, and decreased to the south where

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Raymond J. HofF is Research Plant Geneticist, Intermountain Research
Station, Forest Service, U.S. Department of Agriculture, Moscow, ID 83843,
USA. Susan K. Hagle is Pathologist, Northern Region, Forest Service, U.S.
Department of Agriculture, Missoula, MT 59807, USA. Richard G. Krebill
is Assistant Station Director and Pathologist, Intermountain Research
Station, Forest Service, U.S. Department of Agriculture, Ogden, UT 84401,
U.S.A.

whitebark pine occurs in higher and drier sites. White-
bark pines grovnng below about 45° N. lat. in Idaho and
Montana have much less rust than whitebark of higher
latitudes. However, in the Cascade-Sierra Nevada Moun-
tains the rust has spread to about lat. 36° 15' N., where it

Pinus albicaulis

BH malor subalpine component

Figure 1 — Range map of whitebark pine showing
areas of high, moderate, and little or no blister rust
incidence.

118

has been especially devastating to western white pine
and sugar pine. Whitebark pine too has been heavily
impacted in most of these mountain ranges. But no
infection was found on whitebark pine at Tioga Pass,
Yosemite National Park, at about 39° N. lat. in 1992.

The relative susceptibility of the white pines to blister
rust has been tested by several workers. Bingham
(1972b) siimmarized these reports (table 1). All reports
agreed that whitebark pine was most susceptible of the
tested five-needled pines. Fiu-ther work was done by
Bedwell and Childs (1943) to compare the relative suscep-
tibility of whitebark versus western white pine. These
two pines grow adjacent for much of their ranges, with
whitebark pine at the high-elevation sites and western
white pine lower. They often are foimd on the same site
in the overlap zone. Childs and Bedwell (1948) found that
susceptibility (in terms of nvunbers of cankers and speed
of tree mortality) of whitebark pine was several times
that of western white pine. They concluded that this dif-
ference was due in part to (1) a higher susceptibility of
current year whitebark pine needles and (2) longer needle
retention of whitebark pine (5.3 years for whitebark and
3.8 years for western white pine).

Table 1 — Comparison of susceptibility in white pines to blister rust
by Bingham 1972a or b and a reshuffle by Hoff and others
1980

Species Bingham 1972a or b Hoff and others 1980

Pinus armandii

P. cembra

P. aristata

P. wallichiana

P. koraiensis

P. peuce

P. sibirica

P. parviflora

P. strobiformis

P. strobus

P. flexilis

P. monticola

P. lambertiana

P. albicaulis

119

Bingham (1983) found one western white pine in 10,000
that was free of bhster rust. The possibihty of finding
disease-free whitebark pine trees seemed pretty sHm.

Nonetheless, at sites in northern Idaho and western
Montana where mortahty is over 90 percent due to blister
rust, we have found a few individuals with no or just a
few cankers. Though they certainly are rare (fig. 3). This
gives us hope that there are resistant trees that can be
used to reestablish populations in these areas.

Hofif and others (1980) presented data indicating that
three phenotypically resistant whitebark pine trees in-
cluded in a species trial did indeed have resistance to blis-
ter rust. In this same study, resistant collections of west-
em white pine and sugar pine were also included. This
resulted in a reshuffle of the white pines for susceptibility
to blister rust (table 1). In this comparison, whitebark
pine moved up along with resistant western white pine
and resistant sugar pine. However, these same species
were tested in France by Delatour and Birot (1982), in
Japan by Yokota (1983), and in Germany by Stephan
(1985), and the resistance level of the collections of white-
bark, western white, and sugar pine was much lower
(tables 2, 3, 4). The only explanation is that there are
different races of blister rust at these other test sites.

In spite of this bad news, our plans are to proceed to de-
velop resistance in whitebark pine, including as many

Figure 3 — Blister rust-resistant candidate trees at Gisborne Mountain (A)

120

genes for resistance as possible. If new races of the fungus
appear, the new resistant varieties should have adequate
flexibility to exhibit resistance to the new races. This
seems justified, since high levels of resistance in western
white pine and sugar pine are still being observed in blister
rust resistance tests, even though several races of the fun-
gus are evident.

ACTIVE WORK AND PRELIMINARY
PROGRESS

Blister Rust Surveys of Stands, Clearcuts, and
Bums — Onr objectives are to determine the level of blis-
ter rust damage over the reuige of whitebark pine and to
locate phenotypically resistant trees. As mentioned, the
degree of damage is associated with latitude. Phenotypi-
cally resistant trees have been located in most stands
where high mortality to blister rust has occurred. We
also observed a relatively large amount of natural regen-
eration occurring in clearcuts and bums adjacent to
stands of whitebark pine that have had high rust-caused
mortality. Table 5 shows the number of trees per acre
and the degree of infection by blister rust in four such
stands. Natural regeneration of whitebark pine occurs
from unused seed caches of Clark's nutcracker. The

Burl^e's Peak (B).

Table 2 — Percent of rust-infected seedlings in bulk collections of
white pine species (from Delatour and Birot 1 982)

Geographic

Number

Percent

group

Species

of lots

infected

European

P. cembra

P. peuce

0-20

Asian

P. sibirica

P. parviflora

P. koraiensis

0-9

P. wallichiana

7-67

P. armandii

American

r. ailolala

o
c.

P. albicaulis

40-100

P. flexilis

0-100

P. strobiformis

40-100

P. lambertiana

64-96

P. strobus

88-100

P. monticola

94-100

abundance of regeneration was surprising. It immedi-
ately broxight up questions: At what level of mortality will
there not be enough seed available for caching by the nut-
cracker? If the seed did not come from the adjacent high-
mortality stand, where did it come from?

Artificial Inoculation With Blister Rust — The objec-
tive is to determine the level and variety of mechanisms
that impart resistance in phenotypically resistant trees
and in stands that have had high mortality by blister

Table 3 — Results of inoculation in France and Idaho.

Percent of rust infection on "resistant" crosses
of P. lambertiana and P. monticola from
inoculation in France or Idaho (Delatour and
Birot 1 982)

Seed lot No. France
and cross mean Idaho

Percent

P. lambertiana

41 CF^

41 BF,

41 AF,

P. monticola

43 BF,

43 EF,

43 CF,

43 A F,

43 FF,

43 DF,

47 A F,

46HF2

46EF2

46BF2

46FF2

46G F2

46DF2

46 AFj

46CF2

Table 4 — Comparison of blister rust susceptibility of some white
pines from Stephan (1985)

North America

Percent

Eurasia

Percent

species

infection

species

infection

P. albicaulis

P. armandii

P. aristata

P. cembra

P. balfouriana

P. koraiensis

P. flexilis

P. morrisonicola

P. lambertiana

^97 (76)

P. parviflora

P. monticola

P. peuce

P. strobiformis

P. pumila

P. strobus

100

P. sibirica

P. wallichiana

'Two seedlots were tested; one was not as susceptible as the other.

rust. A trial test was inoculated in August 1991 using
methods outlined by Bingham (1972a). This test included
about 1,000 seedlings from 36 families from 10 populations.
The inoculation was light; only 26 percent of the seedlings
developed needle spots. However, there was considerable
variation among populations (table 6). Even with the light
inoculation, it was encouraging to observe several seedlings
that had what we call "premature needle shed" (McDonald
and Hoff 1970). This is a resistance mechanism in which
needles with needle spots are shed from 9 to 12 months
after inoculation and before the rust fungus has invaded
the stem. This type of resistance has been observed in
several white pine species (Hoff and others 1980). The
seedlings were reinoculated in August 1992 and as of
November 1, 1992, 87 percent had needle spots.

Adaptive Variation of Whitebark Pine — How far

can we transfer seed £ind not cause maladaptation in the
seedlings? The range of whitebark pine extends nearly
19 degrees of latitude and 20 degrees of longitude, and
from 1,524 to 3,354 m. In addition, the species grows in
various ecological situations. It can fill the role of a pio-
neer and a serai. It can exist for long periods as a domi-
nant on harsh sites and even in some stands with the cli-
max species, subalpine fir. Whitebark pine is found in
106 of 225 habitat/phase types in eastern Idaho-western
Wyoming, central Idaho, and Montana (Pfister and others
1977; Steele and others 1981, 1983).

However, whitebark pine is relatively intolerant to
shade and highly sensitive to competition (Arno and
Weaver 1990). Therefore it is frequently restricted to
tougher sites that prevent or restrict the growth of other

Table 5 — Number and infection of whitebark pine seedlings in four
clearcuts

Name of Seedlings Blister rust

clearcut per hectare infection

Percent

Vermillion Pass

445

Upper Coal Creek

1,544

Upper Big Creek

884

Divide Mountain

3,929

121

Table 6 — Variation in various traits of whitebark pine for 1 0 provenances

Preliminary population comparisons

Third-year

riea

hXi

blister

^Ji* Aif An Q n A

1 a*

Lai.

Long.

Elevation

height

stem

DUO

rust

— Percent- -

Cooper Pass

47-32N

115-44W

1,494

17.1

Gisborne

48-21 N

116-44W

1,692

15.0

Freezeout

47-01 N

116-02W

1,707

10.6

Lunch Peak

48-22N

116-22W

1,982

16.9

Brundage Mountain

45-01 N

116-07W

2,195

16.2

Seven Devils

45-21 N

11 6-31 W

2,302

12.8

Saddle Mountain

45-42N

113-59W

2,378

18.1

Porphyry Peak

46-49N

110-44W

2,509

12.0

Boulder Peak

41-35N

1 23-05 W

2,523

10.7

Palmer Mountain

45-04N

110-35W

2,652

16.3

Mean

14.6

species. Because of the combination of several severe en-
vironmental factors, such as short growing season, cold
air, and snow blast (McCaughey and Schmidt 1990),
whitebark pine forms krummholz stands of shrublike
trees at or near timberline. At lower elevations, where
growing conditions are not quite so severe, whitebark pine
grows in nearly pure stands of trees 30 to 90 feet tall
(McCaughey and Schmidt 1990). When environmental
conditions are even less severe, whitebark pine is associ-
ated with other tree species and takes on a form much
like lodgepole pine. When growdng conditions are ideal,
whitebark pine seedlings will grow fairly rapidly (fig. 4).

With such a wide-ranging and ecologically diverse spe-
cies one would expect several geographic races. Since
nutcrackers normally cache the seed at least several him-
dred meters and in many cases several kilometers from
its source, it is also easy to specvilate that there are few
races because genes from several diverse populations are
being continually mixed. Future tests should clarify this
puzzle.

Seed from about 40 stands have been collected. Our
first area of study will be Idaho and Montana. The test
will be established with seed from about 100 populations.
A preliminary test has revealed some interesting data
(table 6). There is considerable variation among the popu-
lations in 3-year-old total height (fig. 4). The tallest seed-
lings were from the Saddle Mountain stand — a population
from high elevation. There is also high variation among
families within stands (table 7). Also, characteristics such
as red coloration of the succulent stem and the prolifera-
tion of lateral buds vary with population (fig. 5). So far,
none of this variation can be associated with geographic
position or elevation.

Inbreeding — In many stands, only a few individuals
remain. This results in much higher opportimity for self-
fertilization (inbreeding). At what level will inbreeding
adversely affect growth and survival of the seedlings?
A study has been initiated to compare growth and other
traits between a highly decimated population and a stand
with very low mortality. Trees at the two sites will be in-
tercrossed, crossed with pollen from the same site, and
compared v^dth wind-pollinated seed.

BREEDING PLANS TO DEVELOP
RESISTANCE

Our surveys have shown that there are many pheno-
typically resistant whitebark pine trees. A traditional
breeding approach could be used to develop a new blister
rust-resistant variety of whitebark pine from these trees.
Three options in this approach will be discussed. How-
ever, because of the high level of natural regeneration of
whitebark pine in clearcuts and burns adjacent to high-
mortality stands, another approach appears more attrac-
tive. This is a more natural method using mass selection
as the genetic selection system, but within a natural set-
ting and with natural selection processes.

The Traditional Approach (fig. 6) — Steps to success
are:

1. Locate blister rust-resistant phenotypes of white-
bark pine. Most stands have a few candidate trees, but
a high number is required so that we can be assured of

Figure 4 — Three-year-old container-grown white-
bark pine.

122

Table 7 — Variation among whitebark pine families within the
Gisborne Stand

'amily

Third-year
neignt

Red stem

Prolific

DUO

Diister
rust

- - Percent —

Mean

finding several genes for resistance. Thus, many stands
will have to be visited. The breeding population for west-
em white pine is over 3,000 phenotypically resistant trees
and is considered a minimum nimiber of parent trees to
assure maintenance of resistance against several races of
the rust fungus.

2. Collect wind-pollinated seed from each tree. In spe-
cial cases, where there are just one or two trees, artificial
pollination by other candidate trees may be advisable.

3. Sow seed, grow seedlings, and inoculate with blister
rust in the fall of their second growth period.

4. Data over the next 4 years, after inoculation, will
reveal the most resistant parents of the phenotypically
resistant candidate trees, along with their resistant
progeny.

Options for developing the new resistant variety are:

1. Selected resistant parents can be grafted into a seed
orchard.

2. Resistant seedlings can be used to establish a seed
orchard.

3. Surviving seedlings from rust tests that were not
used in the seed orchard can be outplanted in a natural
site. The numbers for this planting can probably be in-
creased by rooted cuttings of the surviving seedlings.

Timing: The first five steps cotdd be done in about 10
years. Time to flowering of grafts could be 10 years, but it
will probably take from 40 to 50 years for their progeny to
flower. New technology may shorten this time.

Problems:

1. What is the natural level of resistance?

2. How far can seed be transferred without serious
maladaptation?

Natural Selection Stand — In many localities in Idaho
and Montana whitebark pine has been decimated by blis-
ter rust, leaving just a few individuals per hectare. Many
of these remnant stands are adjacent to burns or clear-
cuts. In at least several dozen cases these openings

contain high nvunbers of whitebark pine seedlings and
saplings (table 5). With careful management we may be
able to use these newly established stands for our piir-
pose. Only three major steps need to be taken:

1. Select the areas within the clearcuts and bxarns that
will be managed for whitebark pine (fig. 7A).

2. Clean the selected area of competing trees and
shrubs (fig. 7B). Whitebark pine is a poor competitor. In
fact it probably can be grown at much lower elevations if
the site is kept free of competition.

3. Let nature take its course concerning the selection of
the most resistant trees (figs. 7C, D).

Problems:

1. What is the natural level of resistance? We need to
know the level of resistance (1, 5, or 10 percent) in stands
to choose sites that have a good chance for success. If only
1 percent of the seedlings are resistant when produced by
a mature stand that has had 90 to 95 percent mortality
by blister rust, a candidate area woidd have to have an
unusually high number of seedlings to end up with a rea-
sonably stocked stand. If 10 to 20 percent are resistant,
the probability of producing a new stand would be consid-
erably better.

Figure 5 — Proliferated lateral buds on whitebark pine.

123

Locate Parents
(Resistant Phenotypes)

Collect Wind or
Cross-Pollinated Seed

Blister Rust Test
1 0 Years

Graft Seed Orchard
Production
1 0 Years

Seedling Seed Orchard
Production
40 Years

Outplant Survivors
From Blister Rust
Test

Stand Establishment
50 Years

Total Time
70 Years

Total Time
100 Years

Total Time
60 Years

Figure 6 — Options for traditional breeding methods for developing new varieties of blister rust-resistant
whitebark pine, including completion times.

2. How much inbreeding is occurring in the decimated
stands? Severe inbreeding might result in reduced
growth and survivabihty.

3. What is the minimimi number of cones £ind seed
needed before seed caches are effective for regeneration?
Some sites may be beyond help.

4. When these sites start producing resistant seed how
far can the seed be transferred?

Timing (fig. 8): First-generation blister rust selection
has already been made on parent trees. Seedling estab-
lishment has already occurred on dozens of sites. Natursd
selection by blister rust is occmring. In 40 to 50 years es-
tablished stands on preferred sites could be producing
cones.

However, it is not only the increase of resistance genes
that is important — although this is a key to the species'
future — it is important to maintain as much of the natu-
ral function of whitebark pine forests as possible. In addi-
tion to the severe genetic impact to whitebark pine itself,
the critical ecological functions of this species have been
threatened by loss of integrity of whitebark pine forests.
Thresholds are a key consideration when attempting to
restore severely altered ecological systems (Friedel 1991;
Laycock 1991). The ecological thresholds for whitebark
pine forests are not well understood and with the loss of
most of the original structure of these forests much of the
integrity has been lost. It will take generations to repro-
duce nearly pure, mature whitebark pine forests. In the
interim, components of the ecological systems most af-
fected by the loss of whitebark pine forests may be irre-
versibly altered.

Time is of the essence, but all is not lost. Over large ar-
eas the first generation is mostly gone, but in many areas

the second generation is there; we must act now to save it.
In other sireas the first generation is largely intact. In
these areas immediate emphasis needs to be placed on re-
taining or restoring the ecological function of whitebark
pine. Protective measures, such as pruning and Ribes
population manipidation may be used.

What we are trying to say in this section is that there is
much that can and must be done immediately to maintain
and restore whitebark pine in the ecosystem.

BLISTER RUST ON WHITEBARK
PINE, A GENETICS PERSPECTIVE

The general level of resistance of the original popula-
tion (before the introduction of blister rust) is less than
1 percent. Therefore, the genetic consequence of the blis-
ter rust epidemic for most of the range of whitebark pine
is to push it through an extremely small bottleneck. In
some areas where the number of surviving trees has been
reduced to only a few over several hundred hectares, the
bottleneck may result in entire populations being lost.
Figure 1 indicates geographic populations of whitebark
pine that will have the most immediate difficulty (high in-
cidence area), those that will have only minor problems
(moderate incidence area), and those that are not likely to
have a problem (little or no incidence). Fiulher research
clarifying levels of natural resistance will be a significant
aid to predicting the adequacy of natural selection in the
various areas. On some sites, time alone may return
whitebark pine to its former levels; on other sites, white-
bark pine may be lost if not aided by genetic improvement
and silvicultural programs.

In areas where there are many surviving trees within a
very reduced popidation, the concern will be the damage

124

Whitebark Pine Competing Species

Figure 7 — Development of a natural selection stand: A, A clearcut before cleaning containing whitebark
pine and competing trees; B, clearcut after cleaning; C, clearcut after infection by blister rust and selection
of most resistant trees; D, the natural selection stand or seed orchard after growth of selected trees.

Locate Sites
Remove Competing
Trees

Stand Establishment
50 Years

Some stands are
already 20 years old

Total Time
50 Years or Less

Figure 8 — Natural selection method for developing
a new vahety of blister rust-resistant whitebark pine,
including completion times.

done by inbreeding. In most conifers inbreeding is detri-
men1:al. Inbred trees generally grow slowly, and detrimen-
tal genes such as for albinism, short leaves, and stunted
seedlings may often occur. However, whitebark pine is of-
ten transported by birds to areas where there is no other
whitebark pine (Linhart and Tomback 1985). Because
there are only a few individuals in the start-up stand, in-
breeding would be high. This may have been the most
common way of past population expansion and, over time,
lethal and other detrimental genes may have dropped out
of the population leaving a species that is not adversely
affected by inbreeding.

Even vTith all these dire genetic consequences, the ge-
netic opportimity for the survival of whitebark pine in the
high-mortality zone appears very high, especially with the
intelHgent help of humans. There are many phenotypi-
cally resistant trees. These surely can be used as a base

125

for new varieties of whitebark pine resistant to blister
rust. But most exciting, the high natural regeneration of
whitebark pine stands in clearcuts and burns — even in ar-
eas where there appear to be insufficient nimibers of par-
ent trees — indicates that natural processes are already
producing a new variety. Further, these trees are grow-
ing well; there do not appear to be inbreeding problems.

CONCLUSIONS

The consequences of the blister rust epidemic have to be
viewed as a disaster for whitebark pine, and a setback for
Clark's nutcrackers, grizzly bears, and other components
of whitebark pine forest ecosystems. Nevertheless, white-
bark pine appears to be equipped with elements that will
permit survival. Most important is seed caching by the
nutcracker even when there is just a small excess. Also,
phenotypically resistant trees seem to be fairly common.
Considering the low numbers of noncankered western
white pine trees together with the much higher suscepti-
bility of whitebark pine, we wonder that there are any
whitebark pine trees at all.

Over large areas, the first generation of whitebark pine
is nearly gone; however, the second generation is often
present. In other areas the first generation is still largely
intact. To save whitebark pine and its ecosystem in the
most sensitive areas it is essential that we act now. Main-
taining sufficient ecosystem function for whitebark pine
while developing greater resistance will be a challenge for
land managers.

REFERENCES

Arno, S. F.; Weaver, T. 1990. Whitebark pine community
types and their patterns on the landscape. In: Schmidt,
W. C; McDonald, K J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and manage-
ment of a high-moimtain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 97-105.

Bedwell, J. L.; Childs, T. W. 1943. Susceptibility of white-
bark pine to blister rust in the Pacific Northwest. Jour-
nal of Forestry. 41: 904-912.

Bingham, R. T. 1972a. Artificial inoculation of large num-
ber of Pinus monticola seedlings with Cronartium
ribicola. In: Bingham, R. T.; Hofi", R. J.; McDonald,
G. I., eds. Biology of rust resistance in forest trees.
Misc. Publ. 1221. Washington, DC: U.S. Department of
Agriculture: 357-372.

Bingham, R. T. 1972b. Taxonomy, crossability, and rela-
tive blister rust resistance of 5-needled white pines. In:
Bingham, R. T.; Hoff, R. J.; McDonald, G. I., eds. Biol-
ogy of rust resistance in forest trees. Misc. Publ. 1221.
Washington, DC: U.S. Department of Agric;ilture:
271-280.

Bingham, R. T. 1983. Blister rust resistant western white
pine for the Inland Empire: the story of the first 25
years of the research and development program. Gen.
Tech. Rep. INT-146. Ogden, UT: U.S. Department of
Agriculture, Forest Service, Intermovmtain Forest and
Range Experiment Station. 45 p.

Childs, T. W.; Bedwell, J. L. 1948. Susceptibility of some
white pine species to Cronartium ribicola in the Pacific
Northwest. Journal of Forestry. 46: 595-599.

Childs, T. W.; Bedwell, J. L.; Englerth, R. H. 1938. Bhster
rust infection of Pinus albicaulis in the Northwest.
Plant Disease Reporter. 22: 139-140.

Delatour, C; Birot, Y. 1982. The international lUFRO ex-
periment on resistance of white pines to blister rust
{Cronartium ribicola): the French trial. In: Heybroek,
H. M.; Stephan, B. R.; Weissenberg, K., eds. Resistance
to diseases and pests in forest trees. Wageningen:
Pudoc: 412-414.

Friedel, M. H. 1991. Range condition and the concept of
thresholds: a viewpoint. Journal of Range Management.
44: 422-426.

Hoff, R. J.; Bingham, R. T.; McDonald, G. I. 1980. Relative
blister rust resistance of white pines. European Journal
of Forest Pathology. 10: 307-316.

Laycock, W. A. 1991. Stable states and thresholds of
range condition on North American rangelands: a view-
point. Journal of Range Management. 44: 427-433.

Linhart, Y. B.; Tomback, D. F. 1985. Seed dispersal by
nutcrackers causes multi-trunk growth form in pines.
Oecologia. 67: 107-110.

McCaughey, W. W.; Schmidt, W. C. 1990. Autecology of
whitebark pine (Pinus albicaulis Engelm.). In: Schmidt,
W. C; McDonald, K J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and manage-
ment of a high-mountain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 85-96.

McDonald, G. I.; Hoff, R. J. 1970. Resistance to Cro-
nartium ribicola in Pinus monticola: early shedding of
infected needles. Res. Note INT-124. Ogden, UT: U.S.
Department of Agriculture, Forest Service, Intermoun-
tain Forest and Range Experiment Station. 8 p.

Pfister, R. D.; Kovalchick, R. L.; Arno, S.; Presby, R. 1977.
Forest habitat types of Montana. Gen. Tech. Rep.
INT-34. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermoimtain Forest and Range
Experiment Station. 174 p.

Spaulding, P. 1911. The blister rust of white pine. Bull.
206. Washington, DC: U.S. Department of Agriculture,
Bureau of Plant Industry. 88 p.

Steele, R.; Cooper, S. V.; Ondov, D. M.; Roberts, D. W.;
Pfister, R. D. 1983. Forest habitat t3rpes of eastern
Idaho-western Wyoming. Gen. Tech. Rep. INT-144.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Forest and Range Experiment Sta-
tion. 122 p.

Steele, R.; Pfister, R. D.; Ryker, R. A.; Kittams, J. A. 1981.
Forest habitat types of central Idaho. Gen. Tech. Rep.
INT-114. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermountain Forest and Range Ex-
periment Station. 138 p.

Stephan, B. R. 1985. Resistance of five-needle pines to
blister rust. Allgemeine Forstzeitschrift. No. 28:
695-697.

Yokota, S. 1983. Resistance of improved Pinas monticola
and some other white pines to the blister rust fungus,
Cronartium ribicola, of Hokkaido, Japan. European
Journal of Forest Pathology. 13: 389-402.

126

PERFORMANCE OF PINUS CEMBRA,
P. PEUCE, AND P. STROBIFORMIS
WITHIN AIR-POLLUTED AREAS

Karel Kahak

Abstract — ^An inventory of the surviving species within the
heavily air-polluted area of the Ore Mountains showed a number
of species of the genus Pinus without symptoms of injury. Three
species of the stone pines are described to demonstrate the im-
portance of the geological past of the species. It relates to the
impact of the catastrophes and volcanic activities and sudden
changes of climate for the species' preadaptation to the toxic
impact of a changing environment. The causes of the resistance
seemed to be connected with the long-term migration {Pinus
cembra, P. peace) and hybrid origin (P. peace, P. strobiformis).
The importance of the life history of the species is unquestioned.

During the years 1956 until 1966 the author (Kahak
1971, 1988, 1991) estabhshed an international collection
of species of gentis Pinus in Plzeh, West Czechslovakia,
in the forest locality of Sofronka (latitude 50° N.).

Various species of pines are represented in Sofronka ei-
ther by patterns of populations that come from different
localities within every species range or from groves of in-
dividual trees from species that originated from locations
that were not sufficiently preadapted for the local condi-
tions of the new environment (Kahak 1988). A research
center was established at the Arboretxim Sofronka in 1963
to investigate evolution of the species included within the
collection. The entire range of provenance experiments
with some of the species was analyzed with respect to the
evolution of their distribution area during postglacial
time.

The air pollution caused by extensive industrial emis-
sions resulted in a mass extinction of the Norway spruce
forests in the Ore Mountains in the vicinity of Plzeh. In
the first stage of the extinction, the dying spruce stands
were cleared, but small groves of individual trees, pre-
dominantly the local and exotic species of pines, survived.
Therefore, the staff of the research center at Sofronka was
invited to join the search for an ecologically harmless
method of reconstructing the extinct forests.

Some species of white pines were discovered whose
needles appeared to be unaffected by emissions. These
species of stone pines were: Swiss stone pine {Pinus
cembra L.), Macedonian pine (Pinus peuce Griseb.), and
southwestern white pine {Pinus strobiformis Engelm.).
Their unique viability had helped them to survive under
the highly unfavorable conditions.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Karel Ka&fik, National Park of Simiava, Vimperk, Czechoslovakia.

SWISS STONE PINE

1. Occurs as individual trees in the entfre emission
area between the town of Hora Sv. Sebestiana and Mount
Klmovec (1,210 m above sea level [a.s.l.]). Some trees,
otherwise injured by snow and icing, show no injury from
emissions. Many of them bear germinant seeds (Mottl
and Prudic 1982).

2. One grove of fertile trees 60 to 70 years of age occurs
in the cemetery arboretiun in the town of Vejprty. Their
dimensions are 12 cm diameter at 1.3 m and 9-11 m
height. They bear germinant seeds.

MACEDONIAN PINE

On the Saxon side of the mountains:

1. Forest district Steinbach, Department 157a2, eleva-
tion 800 m a.s.l. A group of 17 trees, age of 55 years,
fertile, with natural regeneration. Dimensions: Diameter
26-35 cm at 1.3 m and 15 m height.

2. Forest district Schmalzgrube, Department 291a3,
elevation 900 m a.s.l. Twelve trees of the same prove-
nance and age as the latter growing on a peat bog! The
trees are healthy and fertile. Dimensions: Diameter
15-40 cm at 1.3 m and 11-14 m height. Tree tnmks have
been injured by the red deer and humans (chmbing
sptars).

3. Forest district Grumbach, Department 331a4, ele-
vation 825 m a.s.l. Stand area is 0.75 ha and trees are
50 years old (1982) and grow in a mixture with Pinus
contorta ssp. latifolia and the movmtain variant of Pinus
sylvestris L. Average height of the stand is 18 m, diam-
eter is 21-35 cm at 1.3 m. Macedonian pine is the most
vital of the named species.

On the Czech side of the mountains:

1. Locality Bludna, Forest Management Horm' Blatna,
elevation 1,000 m a.s.l. Northern slope stand of approxi-
mately 200 trees, age probably 70 years. Origin tmknown,
but probably from activities of the Forest Research Insti-
tute Mariabnmn just before World War I. Many trees were
apparently broken at the height of 5-7 m, but replacement
crowns are now reaching heights of 15 m. Trees are rela-
tively fertile with natural regeneration present.

SOUTHWESTERN WHITE PINE

1. Locality: Cemetery arboretum in the town of Vejprty
with a heavy emission impact. One tree only. Diameter
36 cm at 1.3 m, height 14 m, and age 50-60 years. Tree is
fertile with germinant seeds in the cones Ijdng on the
moist grass!

127

PROBABLE CAUSES OF RESISTANCE
Autochtonous Species

Ten years of observations and experimental activi-
ties in the 7-year period 1982-89 give us the following
indications:

Variability — ^Among the local forest tree species, there
are different degrees of resistance between individual
trees within the stands. Both autochtonous and alloch-
tonous pine species had been gradually reduced, but there
are some examples without any symptoms of injury. Fir
disappeared on the Czech side of the mountain range
first, but drought rather than emissions may have been
the cause. Norway spruce seems to be extinct.

Catastrophic Selection — Norway spruce mostly dis-
appeared in a relatively short time — about 6 years. How-
ever, after its massive extinction, individually resistant
trees of this species were found within the extinct forests.
This appearance is commonly known under the term
"population bottleneck" (Dobzhanski and others 1976) or
"catastrophic selection" (Lewis 1962; Raven 1964) or "une
evolution acceleree (Bouvarel 1960).

Founder Effect — ^According to Mayr's "founder-effect"
(Mayr 1942, 1965, 1979) after the strong reduction of
Norway spruce the reconstruction of its genetic architec-
ture within the small number of survivors (founder popu-
lation) might follow (Kahak 1988). This might give rise
to a new species population endemic to the environmental
conditions that caused the latter extinction.

Allochtonous Species

Hybridization — The Canadian species of spruce show
a high degree of emission resistance compared to the local
Norway spruce. Their life history, with their intricate
genetic relations, may help explain this orientation
(Heimburger and others 1983; Morgenstern and Farrar
1964). Morphological and genetic evidence indicates that
black spruce is a relatively young species. It is thought to
be of an American origin resulting from the hybridization
of a proto- white spruce from eastern Asia migrating east-
ward with a proto-red spruce of North American origin
migrating westward. Differences in morphological char-
acteristics between black spruce and red spruce in the di-
rection of white spruce would indicate such hybrid origin
(Heimburger 1983).

Directional Selection

The most relevant information came from all the for-
eign species formerly introduced to this mountain range.
An excellent example is the life history of lodgepole pine
as described by Hansen (1942, 1943). This species came
from the Cascade Range territories of North America
where it had been influenced by repeated volcanic activ-
ity. Consequently, many generations of lodgepole pine
were subjected to directional selection through the toxic
air and soil, and lodgepole pine adapted to become the
founding pioneer stage of the ecosystems' evolution.

Migration Preadaptation

The other very important evolutionary factor is migra-
tion. It acts on the development of ecological potentisds
as a consequence of the tolerance to the extreme envi-
ronmental conditions (toxic sites), the environmental
changes, and all the preadaptations that could be ac-
quired only during a long migration route such as those
in the Tertiary — one from eastern Asia to southern
Europe (Mirov 1967) and the other from eastern Asia
to Australia (Mayr 1965).

HISTORY OF SPECIES

Mirov (1967) located the center where the genus Pinus
originated in Eastern Asia, where catastrophes like vol-
canic activity and changes in climate gave all the species
of pines a specific starting point to their principal
adaptation.

"Pines however possessed two characteristics since the
Mezozoic origin of the genus: their xeromorphy and their
ability to endure direct sunlight. These characteristics
may be considered as generally the same throughout the
whole paleobotanical history." (Mirov 1967).

Pinus cembra

"During the Ice Age, the high mountain pines of the
Alps either perished or were preserved in the lower areas
not covered with ice. Pinus cembra found occasionally in
northern Italy is the relic preserved in one of such refugia
(Emberger 1944). With subsequent warming up, it mi-
grated again to the high elevations of the central Alps.
There is, of course, a possibility of the existence of non-
glaciated refugia even in the central parts of the Alps.
Pinus cembra might have survived in some of these few
areas that remained free from the ice." (Mirov 1967).

The indications are that P. cembra and many other spe-
cies of the eastern Asia flora migrated to the Alps during
the Tertiary. During the last glaciation they occurred in
an almost continuous belt of higher elevations. Conse-
quently, because of the enlarged area of its distribution,
P. cembra developed a well-supported degree of genetic
variability. This species of stone pine survived the last
glaciation, and within its refugias there was probably an
increasing frequency of autogamy that must be taken into
account (Bannister 1965). After the retreat of the glacier,
the migrating progenies probably intercrossed at their
contact zone and thus acquired their increasing variabil-
ity as well. Together with the very short period of the
last glaciation, the imdiminished variability still main-
tains itself today. Pinus peuce apparently has a similar
pattern of the life history.

Pinus peuce

According to Mirov, "There are indications that Medi-
terranean pines migrated from eastern Asia along the
mountain ranges, that once extended north of and parallel
to Himalaya." (Mirov 1967).

"Endemic Pinus peuce is very local in Bulgaria, Albania
and Yugoslavia. It is a haploxylon pine closely related to

128

Pinus griffithii of the Himalaya. Before the great migra-
tion, Pinus pence has been crossed with other haploxylon
pines. Judging by the chemical composition of its tiirpen-
tine (Illoff and Mirov 1956) it has a close affinity to east-
em Asiatic and western American pines" (Mirov 1967).

The hybridization should be a source of variation for ad-
aptation to new environments (Lewontin and Birch 1966).

The whole Mediterranean region experienced many tec-
tonic changes throughout the Tertiary. In the Quarter-
nary there occurred intensive faulting and sinking of
considerable areas. This resiilted in the formation of sep-
arate moimtain ranges and many islands. As a conse-
quence, the present distribution of pines in the Mediter-
ranean region is sporadic and irregular (Mirov 1967).

Pinus strohiformis

Southwestern white pine appear to have morphological
indications of a hybrid origin. Critchfield and Little
(1966) said: "This species forms a link both geographi-
cally and morphologically, between the neighbors Pinus
flexilis and Pinus ayacahuite. Southwestern pine is ap-
parently separated from its southern relative Pinus aya-
cahuite by a 100 mile gap that does not seem to corre-
spond to any comparable discontinuity in suitable
habitats."

We have popvilation samples of Pinus strohiformis in
our arboretum that originated from Greenlee, AZ, at lati-
tude 33°39' N., longitude 109°16' W., elevation 2,740 m
a.s.l. This stand is now 30 years of age. Its fertility began
at the age of 22 years. Its cones are surprisingly variable
in size, from the smaller cones similar to Pinus flexilis to
very big cones with apophyses, similar to Pinus aya-
cahuite var. brachyptera. Such a variation supports the
hypothesis about the hybrid origin of this species.

TESTS OF SPECIES POTENTIALS

The response of a species to the impact of an environ-
mental stress allows us to recognize intrinsic characteris-
tics and appearances that we do not usually encounter un-
der normal circumstances. Therefore, we are focused on
testing the collection of pine species at the arboretum
within the zone where there is a heavy impact of both
emissions and severe mountain climates.

CONCLUSIONS

The threatened trees of some forest species imder heavy
emissions often show a surprising ability to endure the se-
lection pressure of the toxic environment. Evidently this
depends on the life history of the species in question.
Some conditions under which we feel the tested species
have a better predisposition to siirvive are:

1. The species is of hybrid origin.

2. The species has imdergone a long-term interconti-
nental migration.

3. The species has been subjected to repeated strong
directional or even catastrophic selection under the im-
pact of volcanic activity. Volcanic areas are noted for
their toxic air and soil similar in some ways to industrial
emission areas.

REFERENCES

Bannister, M. H. 1965. Variation in the breeding system
of Pinus radiata. In: Baker, H. G.; Stebbins, G. L., eds.
The genetics of colonizing species. In: Proc. First Int.
Un. Biol. Sci. Symp. Gen. Biol. Davis; 1964 February.
New York and London: Academic Press: 353-374.

Bigelow, R. S. 1965. Hybrid zones and reproductive isola-
tion. Evolution. 19: 449-508.

Bouvarel, P. 1960. Note sur la resistance au froid de
quelques provenances de pin maitime. Revue forestiere
francaise. 7: 495-508.

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis-
tribution of the pines of the world. Misc. Publ. 991.
Washington, DC: U.S. Department of Agriculture,
Forest Service. 97 p.

Diels. 1910. ex Mirov (1967).

Dobzhanski, Th.; Ayala, F.; Stebbins, L. G.; Valentine,

L. W. 1976. Evolution. San Francisco: Freeman.
Emberger. 1944. ex Mirov (1967).

Hansen, H. 1942. The influence of volcanic eruptions upon

post-Pleistocene forest succession in central Oregon.

American Journal of Botany. 29: 214-219.
Hansen, H. 1943. Paleoecology of the sand dune bogs on

the southern Oregon coast. American Joiirnal of Botany.

30: 335-340.

Heimburger, C. C. 1983. The evolution of Black spruce.

Proc. XDC. Meet. CTIA, Part 2: 163-166. Toronto.
Illoff, P. M., Jr.; Mirov, N. T. 1956. Composition of gimi

turpentines of pines. XXV. A report on two white pines:

Pinus koraiensis from Korea and Pinus peuce from

Macedonia. Amer. Pharm. Ass. J. Sci. Ed. 45: 77-81.
Kahak , K. 1971. Arboretum Sofronka. Introduction

results in 1958-1968. Praha. 38 p. illus.
Kahak, K 1988. Contribution to maintaining continuity of

the Norway spruce in the Ore Moimtains. Folia Mus.

Rer. Natur. Bohem. Occid. Plzai, Botanika 27. 58 p.
Kahak , K 1991. Colonizing ability of some introduced

species of pines. FoKa dendrologica. 18: 187-200. Veda,

Bratislava.

Lewis, H. 1962. Catastrophic selection as a factor in spe-
ciation. Evolution. 16: 257-271.

Lewontin, R. C; Birch, L. C. 1966. Hybridization as a
source of variation for adaptation to new environments.
Evolution. 20: 315-336.

Mayr, E. 1942. Systematics and the origin of species. New
York: Columbia University Press.

Mayr, E. 1965. Summary. In: Baker, H. G.; Stebbins, G.
L., eds. The genetics of colonizing species. In: Proc. First
Int. Un. Biol. Sci. Symp. Gen. Biol. Davis; 1964 Febru-
ary. New York and London: Academic Press: 553-562.

Mayr, E. 1979. Evolution xmd Vielfalt des Lebens.
Springer-Verlag.

Mirov, N. T. 1967. The Genus Pinus. New York: Ronald Press.

Morgenstem, E. K.; Farrar, J. L. 1964. Introgressive
hybridization in red spruce and black spruce. Tech.
Rep. 4. Faculty of Forestry, Univ. of Toronto.

Mottl, J., Prudic, Z. 1982. Analyza riistu limby na
Klmovci — V Krusnych horach a jejf roubavanci.
Tschechisch. Zpravy les. vyzkumu 27(4): 9-13.

Parsons, P. A. 1963. Migration as a factor in natural selec-
tion. Genetica. 33: 184-206.

Raven, P. 1964. Catastrophic selection and edaphic
endism. Evolution. 17: 336-338.

129

COMPETITION AND CROWN
CHARACTERISTICS OF WHITEBARK
PINE FOLLOWING LOGGING IN
MONTANA, U.S.A.

Todd Kipfer
Katherine Hansen
Ward McCaughey

Abstract — This study was designed to evaluate crown develop-
ment of whitebark pine (Pinus albicaulis) subjected to intertree
competition in young mountainous stands in the Yellowstone
area. A distance-dependent competition index was used to assess
intertree competitive effects on crown growth. Significant in-
verse correlations were found between the competition index and
height, crown diameter, crown diameter/height, and crown vol-
ume; however, only 20 percent of the variation was explained by
measured variables. Competition thresholds, indicating when
crown growth becomes limited, were estimated. In future stud-
ies, crown vigor or density measures could perhaps yield a better
competition indicator.

This study addresses the response of regenerating
whitebark pine {Pinus albicaulis) to competition following
logging. Human activities, such as logging, in the movm-
tainous environments of western North America are cre-
ating a variety of impacts at both the individual scale and
at the ecosystem scale. Concern about regeneration
crown growth and future cone production in logged sites
prompted this study. Cone production is a product of
mean shoot production multiplied by the number of fertile
shoots (Weaver and Forcella 1986). Effects of competition
were hypothesized to affect crown development of white-
bark pine. Intertree competition was measured on indi-
vidual, regenerating whitebark pine in logged sites in
southwestern Montana near Yellowstone National Park.
This study is part of a larger research project conducted
by Kipfer (1992).

Whitebark pine are found on sites with a wide range
of geologic, geomorphic, and ecologic conditions (Hansen-
Bristow and others 1990). Their growth characteristics
depend on numerous factors, including competition, that
vary spatially and temporally. Quantifying the influence
of competing vegetation on conifer grov^i^h may help to de-
termine silvicultural prescriptions such as thinning inten-
sity, optimize management decisions, and provide a basis

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Todd Kipfer is a Doctoral Student, Department of Geography, Arizona
State University, Tempe, AZ 85281; Katherine Hansen is an Associate
Professor of Geography, Department of Earth Sciences, Montana State
University, Bozeman, MT 59717; Ward McCaughey is a Research For-
ester, USDA Forest Service, Intermountain Research Station, Forestry
Sciences Laboratory, Bozeman, MT 59717.

for growth models of young forest stands (Wagner and
Radosevich 1991b). For example, one management option
for whitebark pine is to increase or sustain seed produc-
tion for regeneration and for a food source for the Clark's
nutcracker (Nucifraga columbiana), the red squirrel
{Tamiasciurus hudsonicus), the grizzly bear (Ursus
arctos horribilis), and a host of other birds and mammals
(Craighead and others 1982; Kendall 1983; Kendall and
Amo 1990; McCaughey and Schmidt 1990). Seed produc-
tion may increase with v^dder, more diflFuse crowns (Eggers
1986; Spurr and Barnes 1980), and, therefore, silvicultural
prescriptions that increase crown diffiisivity are desirable.
The effects of silvicultural management options for main-
taining or enhancing whitebark pine ecosystems have had
limited study (Eggers 1990; Schmidt and McCaughey
1990).

PREVIOUS STUDIES

Whitebark pine is considered relatively intolerant of com-
petition and shade (Amo and Weaver 1990), although the
trees may be more shade tolerant at an early stage and
less tolerant at later developmental stages (McCaughey
and Schmidt 1990). Although initial observations by
Eggers (1990) indicate that suppressed whitebark pine
seedlings and saplings respond little to the removal of for-
est competition, the effect of competition on postlogging
regeneration needs further study. Weaver and others
(1990) assessed competition in terms of stand develop-
ment and hypothesized that although the density of
whitebark pine seedlings remained approximately con-
stant across stands of different ages, growth from seedling
to sapling size could not be supported in stands older than
100 years. They suggested that clearings needed to be
larger than 10 m^ for seedlings to grow to sapling size.

The effect of competition on the growth of individual
trees has been studied for many economically important
trees in plantations or homogeneous stands (Daniels and
others 1986; Tome and Burkhart 1989; Wagner and
Radosevich 1991a,b). These studies have primarily used
individual tree grovvi:h models based on competition indi-
ces and have shown an inverse relationship between the
amount of competition from neighbors and individual tree
growth performance (Daniels 1976; Daniels and others
1986; Lorimer 1983; Tome and Burkhart 1989). This
study focused on crown development of whitebark pine
as related to a distance-dependent competition index.
Managers could use this index to determine the optimum

130

spacing or number and spatial arrangement of competi-
tors for maximimi crown development of whitebark pine,

METHODS

Three logged stands were sampled in the Gallatin Na-
tional Forest in southwestern Montana near Yellowstone
National Park. The stands ranged in elevation from an
average of 2,260 m at Moose Creek to 2,475 m and 2,535 m
at Teepee Creek 1 and 2, respectively. The three stands
had been clearcut between 1968 and 1972. Data collec-
tion focused upon intra- and intertree characteristics for
whitebark pine and neighboring trees. One hundred
whitebark pine individuals (more than 0.5 m tall) were
identified within each of the three stands by a systematic-
random sampling procedure along transects. Three
different-sized circular plots were nested around each in-
dividual whitebark pine. Within each circular plot, trees
of specific sizes were identified as competitors using the
following criteria: within the 3-m-radius plot, all tree
seedlings and saplings were identified as competitors;
within the 6-m-radius plot all trees with a diameter at
breast height (d.b.h.) greater than or equal to 4 cm and
less than 10 cm were identified as competitors; and within
the 9-m-radius plot all trees with d.b.h. greater than or
equal to 10 cm were identified as competitors. This
method approximated a fixed angle gauge sweep (such as
that used by Tome and Burkhart 1989) and requires that
trees farther away from the individual tree be larger to be
considered competitors.

Five attributes were measured for each whitebark pine:
(1) age, (2) height, (3) diameter at breast height (d.b.h.),
(4) average live crown diameter, and (5) height of the
maximum crown diameter above the ground surface.
Crown and height variables were measured to the nearest
0.1 m and d.b.h. to the nearest 0.1 cm.

Total sample size was reduced from 300 to 220 after
tree ages showed that some selected whitebark pine indi-
cated some had established prior to logging (advance re-
generation). It is difficult to evaluate intertree competi-
tion prior to logging, and these were therefore eliminated
from further study. A ratio of crown diameter to tree
height iCD/H) was computed to reflect the degree of hori-
zontal crown growth. Crown volume was computed using
the cone volume equation: (Volume = 1/3* PI * R'^ * H).
Measurements on competing trees within the 9-m-radius
plot were tree species, total tree height, diameter at
breast height, and distance (measured to nearest 0.1 m)
to the subject whitebark pine.

A distance-weighted size-ratio index or distance-
diameter index (Alemdag 1978; Daniels 1976; Hamilton
1969; Hegyi 1974; Tome and Burkhart 1989) was selected
to develop a competition index, from which intertree com-
petition could be quantified. The distance-diameter index
is a distance-dependent index that svuns a size ratio of
diameters (d.b.h. or basal diameter) of competing trees to
the diameter of a subject tree. The general theory is that
larger competing trees, relative to the individual tree,
contribute more competition than do smaller trees. The
index also assumes that competitive influence decreases
with increasing distance between the individual tree and
the competitor.

The distance-diameter index was modified for white-
bark pine because the trees often had multiple stems or
did not reach breast height. Height was substituted for
diameter and distance weighting functions were applied
to develop the modified competition index:

C.I.= Z (if./i/.)*function(Z)/ST..)
i=l ' '

where C.I. is the competition index; if. is the tree height
of the subject tree; H. is the tree height of the competing
tree; and DIST.^ is a distance decay function. Three dis-
tance decay functions were used: {1/DIST , l/DIST^ , and
e-°'s^y).

Regression analysis was used to evaluate the relation-
ship between competition and crown characteristics of the
whitebark pine. The three distance-decay functions were
used as dependent variables and tree height and three
crown measures (crown diameter, crown diameter to tree
height ratio, and crown volimie) as independent variables.

RESULTS

Competition index values varied according to the form
of the distance-decay function used for whitebark pine.
Index values calculated using 1/distance and 1/distance^
were similar with mean values ranging from 13.12 to
17.12. Index values using the e~^'^^'^^ decay function
ranged from 3.11 to 4.22.

Tree heights of whitebark pine were significantly corre-
lated (P = 0.007) with the competition index; however,
values for the three distance-decay functions (0.033 for
1/distance, 0.027 for l/distance^, and 0.032 for e"^''*^^^)
indicate that the linearity of that relationship is poor as
shown for 1/distance in figure 1. The intercept value in
the regression equation (y = a + bx) was 2.047, which may
represent maximimi tree heights for whitebark pine in
the age classes represented by this study, when growing
in the absence of competition.

Crown diameters of whitebark pine were significantly
correlated {P = 0.000) with the competition index, shown
for 1/distance in figure 1, with R^ values varying depend-
ing on the form of the distance-decay function (0.131 for
1/distance, 0.096 for l/distance^, and 0.128 for e"^*^^^^).
The intercept value (0.820) of the regression equation
poorly represents the maximum possible crown diameter,
since crown diameter values above 1.0 m were common.

The crown diameter to tree height ratio and crown vol-
ume were significantly correlated to competition index as
shown for the 1/distance decay function in figure 1. Re-
gression R^ values for crown diameter to tree height ratio
varied depending on the form of the distance-decay func-
tion in the competition index equation (0.122 for 1/distance,
0.079 for l/distance^, and 0.120 for e-^'^'=^). Similar regres-
sion i?^ values were obtained in correlating competition
index with crown volume using the three forms of the
distance-decay function (0.092 for 1/distance, 0.058 for
l/distance^ and 0.091 for e-^^^-^^).

The 1/distance decay function in the competition index
model yielded slightly higher R ^ values for tree height,
crown diameter, crown diameter to tree height ratio, and
crown voliraie of whitebark pine. The 1/distance decay

131

S.0-
4.5'
4.0

3.5 ■

ao
^6

2.0

1.0

as
ao

Tro halght » 2.047 - 0.0133'C.I.
R-squarod > 0.033
p = 0.007 n > 220

• • • •

£ li *•* t .y . r

— IX)-

0.0

Crown diameter > 0.820 - 0.0117*C.I.
R-tquarad = 0.131
p - 0.000 n - 220

• a • a

• aa «

a a

•aaa a a a a

aaa • a • a a
■ a aa a a
a a a

a a a aa
•a a «a a

a aa •aaa

•aa a aa a«

a a a a

aaa a aa •

a aa>* a

a a aa • a

a aaa

a a
a a

a a a
a a

Competition Index (I/distance)

Competition index (l/dietance)

Crown diameter/Tree height « 0.428 - 0.00481 'CI.
R-squarad ' 0.122
p - 0.000 n 3 220

Sa • • a
a'« a-a a
J a a • ••

aaa a •
a^ ^'V.-.a ;C

a •
• a • • • a

a a a

• ^"'^

•

10 20 30 40 60

Competition index (I/distance)

Crown vokima <s 0.300 - 0.00056*0.1.
R-aquarad • 0.002
p « 0.000 n = 220

# •

• t •

•••'a *.

10 20 30 40 SO

Competition index (1/distance)

Figure 1 — Bivariate scatterplots of total tree heiglit, crown diameter, crown diameter/lielght, and crown
volume of whiitebarl^ pine plotted against competition index using a distance-decay function of 1/distance.
Regression values shown are model and model coefficients, fl-square, p-level, and sample size (n). (Data
from all tiiree study sites — iy/loose Creek, Teepee Creek 1 , and Teepee Creek 2.)

function was used for development of a potential competi-
tion threshold because it was consistently a stronger pre-
dictor of the four growth variables.

Scatterplots were evaluated from each study site for
competition index values above which there was visually
less variability in the data. No identifiable threshold was
observed for whitebark pine in the Moose Creek stand,
and therefore, in order to increase the sample size, the
two Teepee Creek stands were combined. Potential com-
petition thresholds were estimated for total tree height,
crown diameter, crown diameter to tree height ratio, and
crown voltmie for sampled whitebark pine in the two Tee-
pee Creek stands (fig. 2). Variability of data below poten-
tial competition thresholds indicates that tree growth fac-
tors may not be limiting. The majority of whitebark had
competition index values below the estimated threshold
values.

Regression analysis of the competition index against
the four whitebark pine measures using only those trees
greater than the threshold values jdelded poor results for

tree height (R^ = 0.128,p = 0.062, n = 27), crown diameter
(R^ = 0.146, p = 0.106, n = 18), and crown volume (R^ =
0.091, p = 0.120, n = 27). A statistically significant corre-
lation, however, was found between the competition index
and the crown diameter to tree height ratio (CD/H =
0.846 - 0.0142*C.I.: R^ = 0.230, p = 0.000, n = 18).

DISCUSSION

The specific competition index values developed by this
study explain approximately 20 percent of the variation in
competitive pressure exerted on whitebark pine from the
numbers, sizes, and spatial patterns of neighboring trees.
Assessment of competition on crown development of
whitebark pine is difficult when data come from present
spatial relationships and past growing conditions are un-
known. Competition was a significant factor influencing
crown characteristics of whitebark pine in this study. It
is important to note that these results only show that
competition is a significant factor influencing whitebark

132

5.0
4.5
4.0

3.5

2.5
2.0
1.5
1.0
OJ

•

< Potential threshold

•

;i.r J» • t, -

• • •

•• •

* * •

•

••• • •• •

• • •

• •

20 30 40 SO

Competition index (1/distance)

( • ••••

• • • • • I

Potential threshold

10 20 30 40 50

Competition Index (1/dlstance)

Potential threshold

•

\, • • '

• • • • • •

• •

m •
•

•

• • •
m

• •

20 30 40 60

Compstltlon Index (1/dlstance)

n
<

o
>

• • •/

Potential threshold

10 20 30 40 £

Competition index (I/distance)

Figure 2 — Potential competition thresholds for bivariate scatterplots of total tree height, crown diameter,
crown diameter/height, and crown volume of whitebark pine plotted against competition index using a
distance-decay function of 1/distance. (Data from Teepee Creek 1 and Teepee Creek 2 only.)

pine. The mechanisms of this competitive influence are
unknown.

The distance-decay function 1/distance in the competi-
tion index model created an expected inverse relationship
of tree growth measure to index. As distance from the
subject tree increases effects from competition trees de-
crease. Tree heights and crown volumes had the poorest
correlations with the competition index. Poor correlations
with crown volume may be due to sampHng procedures.
In future studies, crown vigor and crown biomass meas-
urements (stem spacing) might be used to weight crown
volume measures, perhaps yielding a better indication of
the effect of competition on growth.

Crown diameters and crown diameter to tree height ra-
tios had good correlations with the competition index.
Still, only 12 to 13 percent of the variation in these meas-
ures was explained by the competition index. The addi-
tion of age and an indicator variable for stand improved
this relationship for the crown diameter measure, increas-
ing the R 2 value to 0.250. It did not, however, yield a sig-
nificant correlation for the crown diameter to tree height

ratio. The crown diameter to tree height ratio may repre-
sent a characteristic crown shape for young whitebark
pine.

Poor correlations between whitebark pine crown charac-
teristics and competition index may be due to factors in-
fluencing individual tree size (local density, plant geno-
type, seed size, emergence time, microhabitat variations,
and unknown historical growing conditions). A competi-
tion index coxdd incorporate a variety of growing condi-
tions over time and the use of incremental growth charac-
teristics of annual growth could provide some improvement
without substantial measurement efforts.

Potential competition thresholds estimated for the two
Teepee Creek sites may be artifacts due to relatively few
trees with high competition values, or they may be actual
values of competitive pressure indicating when growth
becomes limited by competition. Multiple-aged stands
with varying tree densities should be sampled in future
studies. To compare stands of different ages, the age-
independent competition index methods of Lorimer (1983)
may provide a strong starting point. Although regression
analysis can comparatively evaluate the influences of

133

competitive pressure on whitebark pine characteristics,
actual relationships are more complex than the linear re-
gression models might suggest. Nonlinear regression
methods should be explored for evaluating competition
influences.

Timber harvests provide an important but declining
part of our regional economy (Powers 1991), and economic
considerations are an important component of forest man-
agement policies. Timber harvest can cause accelerated
slope failures, erosion, and stream sedimentation from
roads (Marston and Anderson 1991) when improperly
done. Road building in association with logging provides
access to whitebark pine regeneration sites; however, it is
not believed that logging will be a major means for regen-
erating whitebark pine (Arno 1986).

It is imperative to find viable management options for
increasing crown growth and the potential for increased
cone production for whitebark pine because of its impor-
tance to the endangered grizzly and many other wildlife
species in the Yellowstone ecosystem. The importance of
quantifying how competing trees influence crown charac-
teristics of whitebark pine in logged stands comes from a
need to increase our knowledge of how competition affects
crown development of whitebark pine.

ACKNOWLEDGMENT

This research was supported in part by funds provided
by the Intermoxmtain Research Station, Forest Service,
U.S. Department of Agricultiu*e.

REFERENCES

Alemdag, I. S. 1978. Evaluation of some competition indi-
ces for the prediction of diameter increment in planted
white spruce. Inf. Rep. FMR-X-108. Canadian Forest
Service Forest Management Institute. 39 p.

Arno, S. F. 1986. Whitebark pine cone crops — a diminish-
ing source of wildlife food? Western Journal of Applied
Forestry. 1(3): 92-97.

Arno, S. F.; Weaver, T. 1990. Whitebark pine community
types and their patterns on the landscape. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposium on whitebark pine ecosystems: ecology
and management of a high-mountain resource; 1989
October 29-31; Bozeman, MT. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Research Station: 97-105.

Craighead, J. J.; Svunner, J. S.; Scaggs, G. B. 1982. A de-
finitive system for analysis of grizzly bear habitat and
other wilderness resources. Wildlands Inst. Monogr. 1.
Missoula, MT: University of Montana. 279 p.

Daniels, R. F. 1976. Simple competition indices and their
correlation with annual loblolly pine tree growth. Forest
Science. 22: 454-456.

Daniels, R. F.; Burkhart, H. E.; Clason, T. R. 1986. A
comparison of competition measures for predicting
growth of loblolly pine trees. Canadian Journal of For-
est Research. 16: 1230-1237.

Eggers, D. E. 1986. Management of whitebark pine as po-
tential grizzly bear habitat. In: Contreras, Glen P.;
Evans, Keith E., comps. Proceedings — grizzly bear habi-
tat symposiimi; 1985 April 30-May 2; Missoula, MT.
Gen. Tech. Rep. INT-207. Ogden, UT: U.S. Department
of Agriculture, Forest Service, Intermountain Research
Station: 170-175.

Eggers, D. E. 1990. Silvicultiiral management alterna-
tives for whitebark pine. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and maneige-
ment of a high-mountain resource; 1989 October 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 324-328.

Hamilton, G. J. 1969. The dependence of volume incre-
ment of individual trees on dominance, crown dimen-
sions, and competition. Forestry. 42: 133-144.

Hansen-Bristow, K.; Montague, C; Schmid, G. 1990.
Geology, geomorphology, and soils within whitebark
pine ecosystems. In: Schmidt, W3anan C; McDonald,
Kathy J., comps. Proceedings — symposium on
whitebark pine ecosystems: ecology and management of
a high-mountain resource; 1989 October 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 62-71.

Hegyi, F. 1974. A simulation model for managing jack-
pine stands. In: Fries, J., ed. Growth models for tree
and stand simulation. Res. Notes 30. Stockholm: Royal
College of Forestry: 74-90.

Kendall, K. C. 1983. Use of pine nuts by black and grizzly
bears in the Yellowstone area. In: Bears — their biology
and management; International Conference on Bear
Research and Management. Madison, WI: International
Bears Association. 5: 166-173.

Kendall, K. C; Arno, S. F. 1990. Whitebark pine— an
important but endangered wildlife resource. In:
Schmidt, Wyman C; McDonald, Kathy J., comps.
Proceedings — symposiima on whitebark pine ecosys-
tems: ecology and management of a high-mountain
resource; 1989 October 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
264-273.

Kipfer, T. R. 1992. Post-logging stand characteristics and
crown development of whitebark pine (Pinus albicau-
lis). Bozeman, MT: Montana State University, Depart-
ment of Earth Science. 83 p. Thesis.

Lorimer, C. G. 1983. Tests of age-independent competition
indices for individual trees in natural hardwood stands.
Forest Ecology and Management. 6: 343-360.

Marston, R. A.; Anderson, J. E. 1991. Watersheds and
vegetation of the Greater Yellowstone Ecosystem. Con-
servation Biology. 5(3): 338-346.

McCaughey, W. W.; Schmidt, W. C. 1990. Autecology of
whitebark pine. In: Schmidt, Wyman C; McDonald,
Kathy J., comps. Proceedings — symposivun on
whitebark pine ecosystems: ecology and management
of a high-mountain resource; 1989 October 29-31;

134

Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 85-96.

Powers, T. M. 1991. Ecosystem preservation and the
economy in the Greater Yellowstone Area. Conservation
Biology. 5(3): 395-404.

Schmidt, W. C.; McCaughey, W. W. 1990. Whitebark
pine — a subalpine species needing silvicultural atten-
tion. In: Schmidt, Wyman C; McDonald, Kathy J.,
comps. Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-mountain
resource; 1989 October 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
tiu-e, Forest Service, Intermountain Research Station:
373-374.

Spurr, S. H.; Barnes, B. V. 1980. Forest ecology. 3d ed.
New York: John Wiley and Sons. 687 p.

Tome, M.; Burkhart, H. E. 1989. Distance-dependent com-
petition measures for predicting growth of individual
trees. Forest Science. 35(3): 816-831.

Wagner, R. G.; Radosevich, S. R. 1991a. Neighborhood
predictors of interspecific competition in young

Douglas-fir plantations. Canadian Journal of Forest Re-
search. 21: 821-828.
Wagner, R. G.; Radosevich, S. R. 1991b. Interspecific com-
petition indices and other factors influencing the perfor-
mance of Douglas-fir saplings in the Oregon Coast
Range. Canadian Journal of Forest Research. 21: 829-
835.

Weaver, T.; Forcella, F.; Dale, D. 1986. Cone production
in Pinus albicaulis forests. In: Shearer, R. C, comp.
Proceedings — conifer tree seed in the inland mountain
west symposium; 1985 August 5-6; Missoula, MT. Gen.
Tech. Rep. INT-203. Ogden, UT: U.S. Department of
Agriciilture, Forest Service, Intermotmtain Research
Station: 68-76.

Weaver, T.; Forcella, F.; Dale, D. 1990. Stand develop-
ment in whitebark pine woodlands. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposimn on whitebark pine ecosystems: ecology
and management of a high-mountain resource; 1989
October 29-31; Bozeman, MT. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agricvilture, Forest Ser-
vice, Intermountain Research Station: 151-155.

135

FIRE ECOLOGY OF WHITEBARK PINE
FORESTS OF THE NORTHERN ROCKY
MOUNTAINS, U.SA.

Penelope Morgan
Stephen C. Bunting
Robert E. Keane
Stephen F. Amo

Abstract— Fires once occurred at intervals between 30 and 300
years in whitebark pine (Pinus albicaulis) forests in the Northern
Rocky Mountains, U.S.A., but since the early 1900's fewer fires
have occurred, contributing to declining abundance of whitebark
pine. In the absence of fire or other major disturbance, whitebark
pine is replaced by other conifers on most of the upper subalpine
landscape. Whitebark pines often survive low-intensity surface
fires. Large stand-replacement fires also benefit this species by
creating the open, burned sites where regeneration is most
successful.

In the Northern Rocky Mountains of western North
America, whitebark pine (Pinus albicaulis) historically
dominated many upper subalpine forests. These high-
elevation forests usually have poorly developed, rocky
soils and are often located within wilderness or roadless
areas. As a consequence, whitebark pine is seldom har-
vested for forest products, but it is important for scenic,
watershed, and wildlife habitat values.

Whitebark pine dominates middle- and late-successional
stages. In the absence of major disturbance, however,
whitebark pine is eventually replaced by the more shade-
tolerant subalpine fir (Abies lasiocarpa) and Engelmann
spruce (Picea engelmannii) in most of its range in the upper
subalpine forest zone of the Northern Rocky Mountains
(fig. 1). Whitebark pine is a common serai component
of upper subalpine forests found on the Abies lasiocarpa/
Vaccinium scoparium, A. lasiocarpa/Luzula hitchcockii,
and A. lasiocarpa/Arnica cordifolia habitat types (Pfister
and others 1977; Steele and others 1981, 1983; Weaver
and Dale 1974). These sites are cold, with July mean
temperatures averaging 13 to 15 °C (Pfister and others
1977). Although annual precipitation averages 610 to
1,780 mm, summer drought is common (Arno and Hoff
1989; Pfister and others 1977; Weaver and Dale 1974).

Whitebark pine is also found in pure stands on relatively
dry and severe, windswept sites near timberline where it
is the climax tree species (Amo and HofF 1989). It is the
sole climax tree species on Pinus albicaulis habitat types
in Montana, central Idaho, and western Wyoming, and in
southern Canada (Amo and Hoff 1989; Steele and others
1981, 1983). Whitebark pine and subalpine fir are climax
codominants on the Pinus albicaulis-Abies lasiocarpa habi-
tat types where subalpine fir growth is stunted in the severe
microclimate (Pfister and others 1977). Climax whitebark
pine forests are usually open, with smcdl patches of trees
of mixed ages interspersed with meadows (fig. 2). Average
July mean temperatures are 10 to 12 °C with severe sum-
mer droughts and frosts; the annual precipitation of 71 to
153 cm falls mostly as snow (Arno and Hoff 1989; Pfister
and others 1977).

Fires are very important to regeneration and survival of
whitebark pine on sites where it is serai. Whitebark pine
often survives these low-intensity surface fires, which more
easily kill associated conifers (fig. 3). Stand-replacing fires
also benefit whitebark pine, for although all trees are usu-
ally killed, whitebark pine regenerates on burned sites more
successfully than many associated tree species (Tomback

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Penelope Morgan is Associate Professor, Department of Forest Re-
sources, University of Idaho, Moscow, ID 83843; Stephen C. Bunting is
Professor, Department of Range Resources, University of Idaho, Moscow,
ID 83843; Robert E. Keane and Stephen F. Amo are Research Ecologist
and Research Forester, Intermountain Fire Sciences Laboratory, Inter-
mountain Research Station, Forest Service, U.S. Department of Agricul-
ture, P.O. Box 8089, Missoula, MT 59801.

Figure 1 — Subalpine fir and Engelmann spruce
now dominate many stands and whole landscapes
where whitebark pine was historically abundant.
Whitebark pine is declining in abundance even
where blister rust is uncommon, such as on this
site east of Yellowstone National Park in Wyoming.

136

and others 1990) (fig. 4). Large stand-replacing fires are
infi-equent, usually occurring only during windy conditions
after prolonged drought. The fires that burned in and
around Yellowstone National Park in 1988 were spectacu-
lar examples of the large, high-intensity fires that periodi-
cally burn within whitebark pine and adjacent forest types.
These fires burned whitebark pine habitats in a patchy,
stand-replacing manner.

During the 10,000 years since development of forests af-
ter the last glacial retreat, fires have had a major influence
on the structure and composition of forests in the Northern
Rocky Mountains (Arno 1980). Fire occurrence has been
significantly altered by human activity. Native Americans
used fire to manipulate vegetation, for hunting, for commu-
nication, and for other purposes (Pyne 1982). More recently,
humans have affected fire occiurence by purposefully or in-
advertently igniting fires, through fire suppression, and by
grazing domestic livestock, logging, or otherwise altering
the fuels available to burn. Efi"orts to suppress fires have
become increasingly effective since about 1935 (Arno 1980;
Pyne 1982).

Figure 3 — Whitebark pine trees often survive fires and individual trees may survive multiple
fires. Information on fire frequency and effects is derived from dates of fire scars on tree
sections such as this one from a 208-year-old whitebark pine tree in the Shoshone National
Forest in northwestern Wyoming. First scarred by fires when only 12.2 cm in diameter at
the base, it was again scarred 60 years later. It was cut in 1 988, 1 1 0 years later.

Figure 2 — Whitebark pine often is the only tree
that can grow on harsh sites. Here, in the Challis
National Forest in central Idaho, whitebark pine is
the climax tree species.

137

Figure 4 — Young whitebark pine trees regener-
ated successfully following a stand-replacement
fire in northwestern Wyoming. All trees were killed
by the fire, which occurred 55 years before this
photograph was taken in 1 988.

WHITEBARK PINE DECLINE

Whitebark pine has decHned in abundance in major por-
tions of its range (Arno 1986). Once important on 10 to 15
percent of the forested landscape in the Northern Rocky
Mountains (Arno and Hoff 1989), whitebark pine mortality
rates averaged 42 percent over the last 20 years in western
Montana (Keane and Arno 1993). Arno and others (1993)
found that for a 200-ha study area in Montana, the per-
centage of stands with at least 20 percent basal area of
mature whitebark pine declined from 37 percent in 1900
to 20 percent in 1991; 14 percent of the area was dominated
by whitebark pine in 1900; none was so dominated in 1991.

Whitebark pine decline is most pronounced on the more
productive sites where subalpine fir and Engelmann spruce
are highly competitive (Arno 1986; Ciesla and Fumiss 1986;
Keane and Arno 1993; Kendall and Arno 1990). Declining
whitebark pine threatens wildlife habitat because the seeds
of whitebark pine are a valuable food for many birds and
small mammals, including the endangered grizzly bear
(Kendall and Arno 1990).

Recent decline in whitebark pine abundance is linked
to less frequent fires (Keane and Morgan, these proceed-
ings; Keane and others 1990). Fires in whitebark pine for-
ests occurred at mean intervals of 30 to 300 years based on
fire history information derived from fire scars and stand
ages (table 1). Fewer fires in the last 50 to 100 years (Arno
and Hoff 1989; Morgan and Bunting 1990) have resulted
in extensive changes in the composition of forests in the
high-elevation landscapes of the Northern Rocky Mountains,
Composition of subalpine forests has shifted dramatically
toward dominance by subalpine fir and Engelmann spruce
(Keane and others 1993). The decline of whitebark pine
has been further exacerbated by the introduced blister rust
{Cronartium ribicola) and the native mountain pine beetle
(Dendroctonus ponderosae), both of which kill whitebark
pine but not subalpine fir or Engelmann spruce.

STAND-REPLACING FIRES

In conditions of extreme drought lasting more than
2 years, fires ignited by lightning and fanned by high winds
can rapidly spread and kill trees in large patches. These
fires usually bum in other forest types as well, converting
large segments of the landscape to early successional plant
communities. Fires that spread through forests at lower
elevations historically burned into the adjacent whitebark
pine forests (Arno and Hoff 1989). In whitebark pine for-
ests, stand-replacing fires typically spread on the ground
(Lasko 1990). Fires may kill trees by scorching foliage or
by heating the bole or roots to lethal temperatures. Some-
times, crown fires occiu- that bum through the tree crowns,
killing all trees in their paths (Lasko 1990).

Stand-replacing fires provide important opportunities
for whitebark pine to regenerate. Many competing tree
species rely on the wind to disseminate seed. Whitebark
pine has a distinct advantage in regenerating following
extensive disturbances (Tomback and others 1990). The
Clark's nutcracker {Nucifraga columbiana) conmionly trans-
ports seeds several kilometers (Hutchins and Lanner 1982).
These birds prefer open, burned areas for caching seeds
(Tomback and others 1990). Thus, although large fires are
infrequent, they are ecologically important in maintaining
extensive whitebark pine forests on the landscape.

LOW-INTENSITY SURFACE FIRES

Low-intensity surface fires also influence the relative
abundance of whitebark pine on the landscape. Such
fires are more frequent and smaller in extent than stand-
replacing fires. Low-intensity fires generally kill young
whitebark pine and both large and small subalpine fir.
Such fires can result in open, parklike stands of nearly
pure whitebark pine (Arno 1986). Some fires probably
burned as low-intensity surface fires but later became
stand-replacing fires when burning conditions were more
severe. That whitebark pine trees often survive surface
fires is evidenced by the many living trees that have scars
from one or more fires (fig. 1).

Table 1— Fire frequency from whitebark pine forests expressed
as the mean and range (in parentheses) of the years
between fires

Fire frequency

Geographical area and reference

144 (55 to 304)

Bob Marshall Wilderness Complex, north-

western fy/lontana (Keane and others 1 993)

80 (50 to 300)

Bitterroot Mountains, Montana (Arno 1980)

30 to 41 (4 to 78)

1 00- to 300-ha stands where subalpine fir

is climax, Montana (Arno 1986)

29 (13 to 46)

10 stands within 100 ha, northwestern

Wyoming (Morgan and Bunting 1990)

300

Lodgepole pine forests adjacent to but

at lower elevations than whitebark pine.

Yellowstone National Park (Romme 1 982)

138

These low-intensity fires are more common on relatively
dry sites, occurring only where stand structures, fuel ac-
cumulation, and microclimatic conditions are conducive.
Thus, such fires result in many small burned patches, in-
creasing landscape heterogeneity.

Where fires are more fi'equent, they are more likely to be
of low intensity. Morgan and Bimting (1990) documented
very fi'equent low-intensity fires on a relatively dry site
supporting serai whitebark pine in open, parkHke stands
within a 100-ha area in northwestern Wyoming. There the
mean interval between fires was 33 years prior to 1867.
Fires were much more common prior to 1850 than they
have been since then (Morgan and Bunting 1990).

FIRE REGIMES

Through time, most whitebark pine forests experience a
mixture of stand-replacement and low-intensity fires. The
frequency of these types of fires within a given landscape
will vary with landscape complexity and heterogeneity.
Stand-replacing fires are more common diiring regional
droughts, and often burn large patches regardless of fuel
loading or stand condition. Fire behavior and effects are
also influenced by the stand structure and fuel accumula-
tion, which are in part determined by the time since last
bum.

Where whitebark pine is cHmax, fires are infrequent and
generally of low intensity. In whitebark pine krummholz
and ribbon forests, fires are infrequent and of variable in-
tensity. When fires do occ\ir, many trees die and regenera-
tion is very slow. Keane and others (1990) predict where
blister rust infection rates are high, climax whitebark pine
forests will convert to herbaceous or shrub commtmities
following fire.

Stand-replacing fires are more common where whitebark
pine is a serai dominant. Stand-replacing fires become in-
creasingly likely with advancing succession (Fischer and
Clayton 1983; Morgan and Bimting 1990).

SUCCESSION FOLLOWING FIRE

Fire is a key process affecting serai whitebark pine forest
structure sind composition. Successional patterns on sites
where whitebark pine is serai are predictable (fig. 5), but
they are not closely tied to stand age or time since last dis-
tiu-bance (Mattson and Reinhart 1990). The stand struc-
ture and the microsites created vary from fire to fire. Con-
ditions for successful regeneration of tree seedlings are
sporadic, depending on favorable climatic and site condi-
tions. Althoiigh both subalpine fir and whitebark pine
may establish soon after a fire, it may take a half-century
or longer for a forest to develop.

Whitebark pine is one of the first tree species to become
established in abimdance following stand-replacing fires
(Weaver and Dale 1974). As a consequence, it often domi-
nates initially, often for up to 225 years or more (Loope
and Gruell 1973; Morgan and Bimting 1990). Early serai
stands are dominated by whitebark pine seedlings and sap-
lings growing along with a dense herbaceous and shrub
understory. Subalpine fir seedlings are often present,
especially close to parent trees that survived the fire, but
they grow more slowly than whitebark pine trees (Arno

Late Serai
(ABLA, PEN,
PIAL, PICO)

Shrub/herb

Late Mid-Seral
(ABLA, PIAL,
PICO, PEN)

Early Sera!
(PIAL, PICO
ABLA, PEN)

Mid-Serai
(PIAL, PICO,
ABLA, PEN)

^^^^^^

Stand-replacing fire
Return to shrub/herb stage

Low-intensity fire
Many PIAL survive

Figure 5 — Generalized forest succession following
fires on sites where whitebark pine (PIAL) is re-
placed by subalpine fir (ABLA) and Engelmann
spruce (PEN) with advancing succession. Lodge-
pole pine (PICO) is a common associate. Low-
intensity surface fires may occur at any stage
but are most likely in mid-seral stands. Stand-
replacing fires are the norm in late-seral stands.
Tree species listed in order of abundance.
Adapted from Fischer and Clayton (1983).

and Hoff 1989). Whitebark pine seedlings are more abun-
dant than subalpine fir seedlings in large burned areas
because whitebark pine seedlings are dispersed farther by
the Clark's nutcracker. Subalpine fir seeds are dispersed
by the wind. Many standing snags and fallen logs are
present. Most whitebark pine trees do not produce large
numbers of seeds imtil at least age 70; most do not produce
any cones xmtil age 50 (Morgan and Bvmting 1992). With
time, whitebark pine is gradually replaced by subalpine fir
and Engelmann spruce. In late-seral stands, both the tree
canopy and the imderstory are dominated by many subal-
pine fir trees.

MANAGEMENT IMPLICATIONS

Continued decline in whitebark pine abundance threat-
ens to dramatically reduce the availability of seeds for the
many animals that rely on them as a food source (Arno
1986).

Fire exclusion greatly reduces opportimities for regen-
eration of whitebark pine. Cone production is higher in
the stands where whitebark pine is healthy and dominant
(Morgan and Bunting 1992). With blister rust reducing
cone production and killing parent trees (Arno and Hoff
1989), the seed available for tree regeneration is rapidly

139

declining (Amo 1986; Keane and Amo 1993). As cone pro-
duction declines, animals eat more of the seeds, leaving
fewer to regenerate. Whitebark pine cone production de-
clines with advancing succession (Morgan and Bunting
1992) and as infection by blister rust increases (Arno and
Hoff 1989). Given the rapidly declining abundance of white-
bark pine in some regions, we must act quickly to create op-
portunities for individuals that are resistant to blister rust
to regenerate. Although whitebark pine is very sensitive
to blister rust, some individual trees are genetically resis-
tant to infection (Hoff and others 1990). If we do not create
opportunities for those trees to regenerate before their cone
production declines or they die in advancing succession,
opportunities for enhancing natural mechanisms of white-
bark pine recovery will be lost.

Three options are available for improving the health and
productivity of whitebark pine stands that are now domi-
nated by subalpine fir and spruce. One option is creating
forest openings through timber harvest, girdling, or other-
wise killing trees mechanically. Cutting trees to create
openings to encourage regeneration of whitebark pine is
possible but may not be economically feasible. Although
there are often small whitebark pine trees growing in the
imderstory of mixed conifer stands, they are often very old
and of poor vigor and are therefore unlikely to respond
when larger trees are removed.

Another option is the liberal use of prescribed fire to cre-
ate openings for regeneration and to favor whitebark pine
in stands now codominated by subalpine fir. Managers can
purposefully ignite fires or allow lightning fires to burn
under carefully prescribed conditions of weather, fuel, and
location. Historically, fire was the primary natural distur-
bance, and it may be the most practical tool for managing
whitebark pine considering economics, policy, and topo-
graphic limitations for using timber harvest.

A third option, to use a combination of techniques, may
be most successful. Arno and Keane are involved in an ef-
fort to test alternative techniques for perpetuating white-
bark pine in a blister rust-infected area of the Bitterroot
National Forest in Montana.

ACKNOWLEDGMENTS

Research funding was provided through the USDA Forest
Service, Intermountain Research Station. This is Contribu-
tion No. 693 of the Idaho Forest, Wildlife and Range Experi-
ment Station, University of Idaho, Moscow, ID 83843.
We thank Bill Fischer and Elizabeth Reinhardt for their
comments.

REFERENCES

Arno, S. 1980. Forest fire history in the Northern Rockies.
Journal of Forestry. 78(8): 460-465.

Arno, S. 1986. Whitebark pine cone crops — a diminishing
source of wildlife food? Western Journal of Applied For-
estry. 1(3): 92-94.

Arno, S. F.; Hoff, R. J. 1989. Silvics of whitebark pine
{Pinus albicaulis). Gen. Tech. Rep. INT-253. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station, lip.

Arno, S. F.; Reinhardt, E. D.; Scott, J. H. 1993. Forest
structure and landscape patterns in the subalpine lodge-
pole pine t)^es: a procedure for quantifying past and
present conditions. Gen. Tech. Rep. INT-294. Ogden, UT:
U.S. Department of Agriculture, Forest Service,
Intermountain Research Station. 17 p.

Ciesla, W. M.; Furniss, M. M. 1986. Idaho's haimted forests.
American Forests. 81(8): 23-25.

Fischer, W. C; Clayton, B. D. 1983. Fire ecology of Montana
forest habitat types east of the Continental Divide. Gen.
Tech. Rep. INT-141. Ogden, UT: U.S. Department of
Agriculture, Forest Service, Intermountain Forest and
Range Experiment Station. 83 p.

Hoff, R. J.; Bingham, R. T.; McDonald, G. I. 1980. Relative
blister rust resistance of white pines. European Journal
of Forest Pathology. 10: 307-316.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Keane, R. E.; Arno, S. F. 1993. Rapid decline of whitebark
pine in western Montana: evidence from 10-year meas-
urements. Western Journal of Applied Forestry. 8(2):
44-47.

Keane, R. E.; Arno, S. F.; Brown, J. K; Tomback, D. F.
1990. Modelling stand dynamics in whitebark pine
forests. Ecological Modelling. 51: 73-95.

Keane, R. E.; Morgan, P.; Menakis, J. 1993. Landscape
assessment of the decline of Pinus albicaulis in the
Bob Marshall Wilderness Complex, Montana, USA.
Missoula, MT: U.S. Department of Agriculture, Forest
Service, Intermountain Research Station, Intermountain
Fire Sciences Laboratory. Review draft.

Kendall, K C; Arno, S. F. 1990. Whitebark pine — an impor-
tant but endangered wildlife resource. In: Schmidt, W. C;
McDonald, K. J., comps. Proceedings — sjmaposium on
whitebark pine ecosystems: ecology and management of
a high-mountain resource; 1989 March 29-31; Bozeman,
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart-
ment of Agriculture, Forest Service, Intermountain Re-
search Station: 264-273.

Lasko, Richard J. 1990. Fire behavior characteristics and
management implications in whitebark pine ecosystems.
In: Schmidt, W. C.; McDonald, K. J., comps. Proceedings —
sjmiposium on whitebark pine ecosystems: ecology and
management of a high-mountain resource; 1989 March
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden,
UT: U.S. Department of Agriculture, Forest Service, In-
termountain Research Station: 319-323.

Loope, L. L.; Gruell, G. E. 1973. The ecological role of fire
in the Jackson Hole area, northwestern Wyoming. Qua-
ternary Research. 3: 425-443.

Mattson, D. J.; Reinhart, D. P. 1990. Whitebark pine on the
Mount Washburn Massif, Yellowstone National Park. In:
Schmidt, W. C; McDonald, K. J., comps. Proceedings-
symposium on whitebark pine ecosystems: ecology and
management of a high-mountain resource; 1989 March
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden,
UT: U.S. Department of Agriculture, Forest Service, In-
termountain Research Station: 106-117.

Morgan, P.; Bunting, S. C. 1990. Fire effects in whitebark
pine forests. In: Schmidt, W. C; McDonald, K. J., comps.

140

Proceedings — symposium on whitebark pine ecosystems:
ecology and management of a high-momitain resource;
1989 March 29-31; Bozeman, MT. Gen. Tech. Rep.
INT-270. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermoimtain Research Station: 166-170.

Morgan, P.; Bunting, S. C. 1992. Cone crops in whitebark
pine (Pinus albicaulis) forests in the Greater Yellowstone
Area. Moscow, ID: University of Idaho, College of For-
estry, Wildlife and Range Sciences. 21 p. Review draft.

Pfister, R. D.; Kovalchik, B. L.; Arno, S. F.; Presby, R. C.
1977. Forest habitat types of Montana. Gen. Tech. Rep.
INT-34. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermountain Forest and Range Experi-
ment Station. 174 p.

Pyne, S. J. 1982. Fire in America. Princeton, NJ: Princeton
University Press. 654 p.

Romme, W. H. 1982. Fire and landscape diversity in subal-
pine forests of Yellowstone National Park. Ecological
Monographs. 52: 199-221.

Steele, R.; Pfister, R. D.; Ryker, R. A.; Kittams, J. A. 1981.
Forest habitat types of central Idaho. Gen. Tech. Rep.

INT-114. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermountain Forest and Range Ex-
periment Station. 138 p.

Steele, R.; Cooper, S. V.; Ondov, D. M.; Roberts, D.;

Pfister, R. D. 1983. Forest habitat types of eastern Idaho-
western Wyoming. Gen. Tech. Rep. INT- 144. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Forest and Range Experiment Station. 122 p.

Tomback, D. F.; Hoffman, L. A.; Sund, S. K. 1990. Coevo-
lution of whitebark pine and nutcrackers: implications
for forest regeneration. In: Schmidt, W. C; McDonald,
K. J., comps. Proceedings — symposium on whitebark
pine ecosystems: ecology and management of a high-
mountain resource; 1989 March 29-31; Bozeman, MT.
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department
of Agriculture, Forest Service, Intermountain Research
Station: 118-130.

Weaver, T.; Dale, D. 1974. Pinus albicaulis in central
Montana: environment, vegetation and production.
American Midland NaturaHst. 92: 222-230.

141

VEGETATION DISTRIBUTION AND
PRODUCTION IN ROCKY MOUNTAIN
CLIMATES— WITH EMPHASIS ON
WHITEBARK PINE

Tad Weaver

Abstract — The distribution and production of vegetation on the
altitudinal gradient (grassland-forest-alpine) was plotted against
climatic parameters to evaluate hypothetical controlling factors.
(1) Whitebark pine (Pinus albicaulis) is likely excluded from
higher zones by a cool growing season or wind-induced drought.
It is probably not excluded by low temperatures occurring during
its hardening, hard, or dehardening seasons. (2) While the lower
physiological limit of whitebark pine is probably set by drought,
its lower realized limit is directly set by subalpine fir (Abies lasio-
carpa) and lodgepole pine (Pinus contorta) competitors and indi-
rectly set by factors that control their distribution. (3) The upper
limits for most other dominant species are probably set by grow-
ing season temperature. The lower Umits are likely set by com-
petition down to the cedar-hemlock (Thuja plicata I Tsuga hetero-
phylld) zone and by drought in drier areas. (4) Production is
strongly correlated (r^ = 0.86) with growing season length (soil
thawed season minus dry soil days). Multiplying season length
by average temperature did not improve the growing season pre-
dictor, perhaps because vegetation at each altitude is especially
adapted to temperatures in its zone.

Vegetation composition and structure vary along gradi-
ents of temperature and precipitation v^^hether the condi-
tion changes with altitude (Daubenmire 1956; Gams 1931;
Weaver 1980) or geography (Holdridge 1967; Walter 1973;
Whittaker 1975). Graphical devices based on simple pa-
rameters such as average annual temperature and total
annual precipitation (Holdridge 1967; Whittaker 1975)
are workable predictors of the vegetation growing in par-
ticular climates. It seems obvious, however, that these
devices succeed, not because the climatic parameters used
are causal, but because the parameters are correlated
with causal climatic factors.

Production of vegetation is largely determined by cli-
mate and is therefore expected to be highest where tem-
perature, moisture, and nutrients are simidtaneously
favorable. In Rocky Mountain vegetation one therefore
expects production to be highest in low-altitude (warm)
moist forests (Thuja-Tsuga) and to decline both with in-
creasing altitude (because of cooling and reduced nutri-
ent availability) (Weaver 1979) and with decreasing alti-
tude (because of decreasing water availability). While it is

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Tad Weaver is Plant Ecologist, Biology Department, Montana State
University, Bozeman, MT 59717.

rarely done, vegetation production should also be success-
fully plotted in a Hutchinsonian (1958) hyperspace with
axes that are either physiologically meaningful or surro-
gates well correlated with physiologically meaningful axes.

The object of this paper is to correlate vegetation
performance — survival and production — ^with physiologi-
cally meaningftd aspects of temperature and moisture.
Study of the distribution of species on these prestunptively
causal axes will eliminate some hypotheses of cause and
sharpen others for experimental tests. While whitebark
pine (Pinus albicaulis) is the primary subject of this paper,
the approach could be applied to other species, as well.

METHODS

Impacts of climatic factors on vegetation distribution
were evaluated by comparing factor levels among seg-
ments on a vegetational gradient. In essence, the method
allows one to deduce (1) that a presmnptive factor that
does not vary between vegetation t3T)es is not controlling
the vegetation changes observed, and (2) that a presimip-
tive factor that does change between vegetation types de-
serves further discussion as directly controlhng, indirectly
controlling, or a noncausal correlate. The method is illus-
trated here with discussion of whitebark pine, but it could
be applied to many other species as well.

The Data

Environmental zones were identified by climax vegeta-
tion occupying them (Daubenmire 1943; Daubenmire and
Daubenmire 1968; Huschle and Hironaka 1980). They in-
cluded, from low to high altitude, desert shrub, dry grass-
land, warm forest, cool forest, and alpine ecosystems.
Specific zones are listed with their Kuchler (1964) type
numbers and representative stations in table 1.

Climates of the environmental zones were characterized
by using climatic statistics from approximately five sta-
tions in each environmental-vegetation zone. To maintain
the integrity of the data, raw data are plotted wherever
possible. When necessary, medians were used to repre-
sent "typical" sites; use of medians deemphasizes sites
vrith conditions intermediate between modal conditions
of adjacent types. The stations chosen include most avail-
able stations and may thus be considered as a "complete
sample." The sample may not be entirely random because
weather stations must be accessible; in higher types, for
example, the sites may be relatively low.

142

Table 1 — Weather stations representative of major Northern Rocky Mountain environmental zones. ^ Zones are listed from high to low altitude

Zone Station locations

Alpine^

Whitebark pine^
Subalpine fir
Douglas-fir
Cedar-hemlock
Ponderosa pine

Idaho fescue
Bluebunch
wheatgrass
Grama grass
Atriplex spp.

White Mountain, CA; Niwot Ridge, CO

Old Glory, BC; Kings Hill, MT; Ellery Lake, CA; Crater Lake, OR

High: Summit, MT; West Yellowstone, MT; Lake Yellowstone, MT; Low: Burke, ID; Seeley Lake, MT; Hungry Horse, MT
East: Hebgen Lake, MT; Lakeview, MT; Lamar, WY; Palisades Dam, ID; Dixie, ID; West: Libby, MT; Lincoln, MT
West Glacier, MT; Sandpoint, ID; Pierce, ID; Avery, ID; Priest, ID

West: Garden Valley, ID; Poison, MT; Potlatch, ID; Kooskia, ID; East: Melstone, MT; Busby, MT; Colstrip, MT; Lame
Deer, MT; Roundup, MT

Gallatin Gateway, MT; Virginia City, MT; Mystic Lake, MT; Wisdom, MT; Bozeman MSU, MT; White Sulphur Springs, MT

Browning, MT; Belgrade, MT; Kalispell, MT; Ennis, MT; Dillon, MT; Augusta, MT; Crow Agency, MT; Billings, MT
Ekalaka, MT; Jordan, MT; Malta, MT; Rock Springs, MT; Rapelje, MT; Chester, MT; Great Falls, MT; Fairfield, MT
Lovell, WY; Worland, WY; Basin, WY; Powell, WY; Deaver, WY

'The vegetation types, with their Kuchler (1964) type numbers (KTXX), are: Deschampsia caespitosa-Carex (KT52), Pinus albicaulis (KT15), Abies lasiocarpa
(KT1 5), Pseudotsuga menziesii (KT1 2), Thuja plicataJTsuga heterophylla (KT2), Pinus ponderosa (KT1 1,16), Festuca idatioensis (KT63), Agropyron spicatum
(KT63), Bouteloua gracilis (KT64), Atripiex (KT40). Abies types from higher and lower altitudes, Pseudotsuga types from east and west of the mountains, and pon-
derosa pine types from east and west of the Rockies are separately presented to emphasize possible differences.

n"he White Mountain alpine site (1 0,1 50 ft) is unusually dry. The Ellery Lake whitebark pine site has only precipitation data; temperature data were therefore
taken from a similar altitude at a nearby White Mountain site (9,645 ft).

Killing Temperatures

The hypothesis that extreme temperatiires might ex-
clude a plant (for example, whitebark pine) from an adja-
cent zone was tested by determining whether the condi-
tion is more extreme in the iinoccupied than in the
occupied zone; if it is not, the hypothesized factor is con-
sidered unlikely to be lethal. This test was used on tem-
peratures of early fall frosts, midwinter extreme lows,
midwinter average lows, late fall frosts, midsummer aver-
age highs, and midsummer extreme highs. Midwinter
lows £ind midsvmimer highs were tested because these
temperatures might exceed the potential tolerances of a
plant; fall and spring frosts were considered because they
might kill partially hardened plants. Fall frost tempera-
tiu-es were measured as the absolute low in the month in
which average temperature first fell below 0 °C; where no
month had an average temperatvire of 0 °C, the values
were interpolated to 0 °C from adjacent months. Spring
frost temperatiu-es were estimated similarly for the
month in which average temperatures first rose above
0 °C. I argue later (see growing season) that average tem-
peratures below 0 °C restrict water and nutrient uptake
and therefore open and close the growing season; the use
of 0 °C is in contrast to the 5 °C used by many phenolo-
gists (for example, Chang 1968). "Absolute" temperatures
were the lowest (or highest) seen in the period of record —
10 years or more.

Starvation Temperatiu*es

The plant might be excluded from a zone free from kill-
ing events if (temperatiire) conditions failed to support
net photosynthesis on an annual basis. Two indices of
growing season temperature were used to test the possi-
bility of such starvation. First, temperatures were aver-
aged across all growing season (defined later) months.
Despite its common use in predicting growth (Chang 1968;
Larcher 1975), this index is expected to underestimate the

benefits of warmer temperatvires because chemical reaction
rates increase exponentially with increasing temperature.
An alternate index of temperature was therefore tested.
In the second index temperatures were replaced with
growth support units (gsu) and these were averaged
across the season. On the assimiptions that Q^^ = 2, that
very slow growth begins at 1 °C, and that native plants
tolerate normal high temperatures of their zones, the
growth support unit curve was constructed by interpolat-
ing between 0 growth rate imits (gsu) at 0 °C, 1 gsu at
1 °C, 2 gsu at 11 °C, 4 gsu at 21 °C, 8 gsu at 31 °C, and
16 gsu at 41 °C. That is, we expect no growth at 0 °C or
below, 1 vmit at 1 °C, 2 imits at 11 °C, 4 units at 21 °C, etc.

Killing Drought

Evaporation equals roughly 2 mm/°C x month
(Daubenmire 1956; Nielsen 1986; Stephenson 1990;
Thornthwaite 1948; Walter 1973). As a result, one might
index drought duration by counting months when average
precipitation (mm/2) is less than average temperature
(C) or drought intensity by summing, across growing sea-
son months, precipitation deficits below the calculated
balance.

A covmt of months registering drought on the Walter in-
dex may overestimate drought duration where the deficit
is small because drought may open late or close early in
the month. To minimize this effect I have normalized
drought duration from a scatter diagram of duration
against deficit: 0-5 mm = 0 months, 6-15 mm = 1 month,
16-35 mm = 2 months, 36-50 mm = 3 months, and
51-85 mm = 4 months.

I have used the simi of warm-season (average tempera-
ture above 0 °C) water deficits as an index of drought
magnitude. I do so with the recognition that, due to ex-
clusion of important factors (for example, wind, Penman
1949, and exponential temperature effects, Stephenson
1990), this index usually underestimates water deficit.

143

Water Starvation

RESULTS AND DISCUSSION

"Water starvation" should occur where plants are not
desiccated, but where water supplies do not support net
photosynthesis on an annual basis. For example, starva-
tion would occur if water were continually available, but
evapotranspiration equaled or exceeded uptake rates. Such
starvation would select for organisms with lower ET rates:
Mesophytic leaves would be replaced by xeroph)i;ic leaves
and leaf exposiu-e might be reduced by reduction of stature
even to the point where plants existed only at the groimd
surface or under transparent rock.

Growing Season

For sensitive introduced plants, frost-free season is of-
ten recorded as an indicator of season length (Burke and
others 1976; USGS 1971).

Since cool region plants normally tolerate growing sea-
son frosts, I argue for a different index. Growth requires
both leaf activity (photos3nithesis) and root activity
(water and nutrient absorbtion). For plants with frost-
insensitive leaves (Burke and others 1976), midday tem-
peratures should be high enough for photosynthesis long
before soils are thawed. Root activity extends (maxi-
mally) from spring soil thaw to the fall refreeze, so this
period should be a better index of growing season length.
Thus I open and close the growing season with 0 °C aver-
age monthly air temperatures. So represented, initial ac-
tivity in grasslands begins at spring soil temperatures
(0-25 cm) near 0 °C and activity ceases in the fall with soil
temperatures as high as 5 °C (Weaver, in preparation).
In, or adjacent to, the subalpine fir {Abies lasiocarpa) zone
the growing season may exceed this period because deep
snow cover may prevent the freezing of soil water.

While the temperature-bounded season probably ap-
plies at high-altitude sites, at low-altitude sites soil dry-
ing may close the growing season before frost does. Thus,
one may better index growing season as warm season
months minus any drought months included in the warm
season. While the Northern Rocky Mountain drought sea-
son comes at summer's end (Weaver 1980), this index will
also apply to regions where the warm season opens with
dry months. It is conceivable that other periodic factors,
for example nutrients or pest attack, might affect growing
season length.

Production

Production estimates for each series were drawn from
the literature (table 2). These are correlated with climatic
conditions that might control them: warm season length,
warm moist season length, and the product of warm soil
season length and temperature. For the latter, two tem-
perature expressions were tested: average temperature
minus 5 °C and Q (the Q^^-based average). The third
(product) indices were expected to be best because they
assume that plants grow when soils are unfrozen-moist
and that their growth rate is proportional to temperature
(over 5 °C for the first linear index and over 1 °C for the
second exponential index).

In the Northern Rocky Motmtains vegetation changes
from grass-shrub (Atriplex, Bouteloua, Agropyron, and
Festuca) to conifer {Pinus ponderosa, Thuja-Tsuga,
Pseudotsuga, and Abies) to alpine graminoid with increas-
ing altitude. Subzones listed parenthetically are described
by Daubenmire (1943) and mapped by Kuchler (1964).
This paper compsu-es weather data gathered in these zones
to test h3TJothesized controls of distribution and production.
The method is illustrated with whitebark pine.

Upward Limits

I speculate that the upward limits of whitebark pine
might be determined by lethal factors (winter low tem-
peratures, spring-fall fi*osts, or desiccation) or production
deficits (starvation) due to inadequate water or heat
units. These h3T)otheses are considered in the following
paragraphs.

If winter lows are lower in the alpine than in the pine
zone below, winter cold might be the excluding factor, oth-
erwise not. Contrary to my expectation, neither absolute
lows nor average January minima are lower in the alpine
than in the pine (or most other) vegetation zone(s) (fig. 1).
The failure of low temperatures to develop at higher alti-
tudes may be due to the high density of cold air: Cold air
entering from the north stays low like mercury poured \m-
der water and air cooled by exposure to cold alpine ground
runs off (Geiger 1965). Since extreme temperatures are
not lower at high than low altitude, I deduce that neither
whitebark pine, nor most other native dominants, are ex-
cluded from high sites by extreme low temperatures of win-
ter. An exception to this generalization seems to appear in
high- or low-altitude fi-ost pockets, where accumulating
cold air may kill trees in winter or spring (Weaver 1990).

For natives of an area, I suggest that frost damage is
more likely when plants are partially hardened (fall) or
partially dehardened (spring) than at midwinter. I expect
(1) winter to be delimited approximately by the fall month
in which surface soils fall below 0 °C and the spring
month in which they rise above 0 °C, (2) winter to be a
season of low root activity, low water, and low nutrient
uptake, and (3) 0 °C soil temperatures to be approxi-
mately coincident with 0 °C air temperatures (Weaver, in
preparation). I therefore (1) compare, across vegetation
types, observed temperature lows in these spring-fall sea-
sons of incomplete frost hardness (fig. 2), (2) discover no
difference, and (3) conclude that frosts probably do not
partition vegetation native to the region. The preceding
discussion depends on the assumption that plants harden
and deharden in phase with air/soil temperatures. This
argument would be fallacious if day length were the pri-
mary controller of the hardening/dehardening process.
However, since both temperature and correlated day
length triggers are important (Salisbury and Ross 1992),
I expect frosts to exclude plants from environmentally dis-
tant, but not environmentally adjacent, vegetation types.

While frosts and winter freezes probably do not exclude
whitebark pine from the alpine (or other plants from the
vegetation zone immediately above), a lack of heat imits

144

Table 2 — Productivity in grams per square meter per year and aboveground standing crop (Abvegr std crp) of major Rocky Mountain vegetation
types

Vegetation type

Productivity^

AKuonr QtH f^m

g/m^/yr

t/ha (age.yr)

Alpine

135

1.35

Thilenlus and others 1974

100-200

1.00-2.00

Scott and Billings 1 964

Pinus albicaulis

'60

^140 ^300-500^

VVCdVd Cll IVJ L^ClIC' 1 / H

'25-75 (53)

PfiQtpr anH nthorQ 1Q77

r IIOLC7I al IVJ \JlllC7IO 1*7/ 1

200-700

350 (300-500)

Forcella and Weaver 1 977

Abies lasiocarpa

1 00-200

160 f300-700)

Aoipt and other*; 1QfiQ

860

357 M06 vr)

Whitt^^kpr ^r\c\ Niprinn 1 07*^

V V 1 II llCir\C7l ai \\a I NIOI ll im I ^ r w

'95-180 (137)

Pfister and others 1 977

100-250 (Old)

Weaver and Forcella 1977

150-250

Landis and Mogren 1975

Ps6udotsuga menziesii

1,550

438 (252)

Whittakpr and Niprinn 1Q7^

'30-170 (150)

Pfi^tpr and nthpr^ 1Q77

100-350 (350+)

Weaver and Forcella 1 977

Thuja plicata-

Tsuaa hsteroohvlla

870

Hanlfiv 1 Q7fi
rial Moy i 7 / u

550

316 (250)

Hanlev 1976

1,380

504 (103)

Hanley 1976

'150-330 (240)

Pfister and others 1 977

Pinus pondorosa

'188

(^larpv/ ^inH Athorc 1Q7^

'30-150 (188)

Pfister and others 1 975

490-570

150-250 (150)

Whittaker and Niering 1975

50-250 (350+)

Weaver and Forcella 1 977

Festuca idahioensis

235

2.35

Collins and Weaver 1978

195

1.95

Weaver and Collins 1977

147

1.47

Daubenmire 1970

Festuca scabrella

152

1.52

Willms and others 1986

Bouteioua gracilis

103

1.03

Weaver 1 983

53-95

0.5-1.0

Hunt and others 1 988

Atriplex spp.

West 1983

'Since merchantable production and standing crop are emphasized in these studies, one might expect the figures to be 50 to 66 percent of those reported in
studies of total production (Weaver and Forcella 1977).

Where production and standing crop are expressed volumetrically, masses were calculated using specific gravities of ABLA=0.38, PIAL=0.40, PIEL=0.35,
PSME=0.45, PIPO=0.43, PICO=0.38, THPL=0.33, TSHE=0.41 (U.S. FPL 1974).

(starvation) may exclude, from an altitudinal zone, plants
of lower zones. Average growing season temperatures in
the alpine (and for most zones below it) are distinctly
lower than temperatures in the vegetation zone below
(fig. 3). This correlation suggests the possibility of — but
does not prove — significant effects of growing season
temperature. In support of this h3T)othesis, production-
temperature relations (discussed below) suggest that the
vegetation of each altitudinal zone is especially adapted
to temperatures occurring in its zone. In this regard, it
is satisfying to see that optimum temperatures for white-
bark pine photosynthesis (20-25 °C, Jacobs and Weaver
1990) are similar to summer maximum temperatures
(10-25 °C, Weaver 1990).

Graphs of estimated soil water availability against veg-
etation type (fig. 4) indicate little or no stress in alpine,
whitebark, subalpine fir, and Douglas-fir zones. On this
basis one might exclude droiight as a factor determining
whitebark's upslope limit. Due to two modifying factors,

this conclusion seems premature. First, as one moves
upslope from forest to alpine (or moimtain meadow),
winds increase and, with increasing wind, drought in-
creases due to reduced water availability — snow may be
blown off site (Daubenmire 1981) and water in uninsu-
lated soils freezes — and increasing water loss — through
abrasion of cuticle (Hadley and Smith 1987) and thinning
of the leafs boundary layer (Gates and Papian 1971; Nobel
1983). I see winter wind effects as probable controllers be-
cause trees invading (or planted into) mountain meadows
are more often desiccated in winter than stunmer. The
wind effect hypothesis is consistent with the fact that
groups of trees sometimes invade sites that individuals
cannot invade alone (Armand 1992; Tranquillini 1979). In
such situations adjacent trees are more likely sheltering
each other from drying wind than from lethal low tempera-
tures. Second, as one moves upslope the average soil be-
comes progressively better drained (Weaver 1979), and
thus effective precipitation is a smaller fraction of total

145

10 20 30

ECOSYSTEM TYPE

Figure 1 — Midwinter low temperatures in 10 Rocky Mountain ecosystems. Low temperatures (°C
and °F) are represented by absolute winter minimums (A, ABSMN) and average January minima
(B, AVJMN). Ecosystem types range altitudinally from alpine (2), down through forests (8-20) to
grass and shrublands (26-36); specifically, they are alpine (2), whitebark pine (8), subalpine fir (10),
Douglas-fir (16), cedar-hemlock (18), ponderosa pine (20), Idaho fescue (26), bluebunch wheatgrass
(28), gramagrass (30), and desert shrub (36). Large and small circles represent data gathered east
and west of the Rockies, respectively. Lines connect median values.

30 40' '0 10

ECOSYSTEM TYPE

Figure 2— The coldest early spring (A, SPR-FRST) and late fall (8, FALL-FRST) frosts (°C
and "F) in 10 Rocky Mountain ecosystems. Ecosystems, symbols, and lines are as in figure 1.

O 10 20 30 40 O 10 20 30 40

ECOSYSTEM TYPE

Figure 3 — Average growing season temperatures in 1 0 Rocky Mountain ecosystems. The first graph
(A, TEMP gs) gives the average growing season temperature (°C and "F). Q gs in the second graph
(B) considers the exponential response of physiological processes to increasing temperature (Q,q).
Ecosystem types, symbols, and lines are as in figure 1 .

146

precipitation at higher than lower altitudes; that is, the
Walter drought indices presented in figure 4 probably over-
estimate water availability on (high-altitude) thin-soil sites.

Downward Limits

I speculate that downward limits of whitebark pine
might be determined by heat, drought, or competition.
The following paragraphs consider these hypotheses.

As one moves downslope maximum temperatures
(fig. 5) rise. That this correlate of tree disappearance is
not controlling is suggested by three facts. Summer aver-
ages in vegetation zones below whitebark pine (10-15 °C,

fig. 3) are well below the whitebark pine optimimi
(20-25 °C, Jacobs and Weaver 1990), so downward migra-
tion might actually improve production. July maxima
(20-27 °C, fig. 5) in zones below are near the whitebark
pine optimvuu temperature (20-25 °C, Jacobs and Weaver
1990) and thus should not be damaging. Long-term
maxima (30-42 °C, fig. 5) in zones below do not eliminate
net photosynthesis (Jacobs and Weaver 1990) and thus
are probably not lethal. In addition, if whitebark pine
trees are well watered, they grow well as lawn trees in the
Agropyron spicatum zone (for example, Belgrade, MT)
where temperatures are far higher than those fovmd on
sites whitebark naturally occupies.

ECOSYSTEM TYPE

Figure 4 — Drought severity in 10 Rocky Mountain ecosystems. The first graph (A)
gives the number of dry months (DRT-MNTHS). The second (B) indexes the an-
nual water deficiency (W-DFCT, mm). Both were calculated after Walter (1973).
The ecosystem types, symbols, and lines are as in figure 1 .

0 10 20 30 40 0 10 20 30 40

ECOSYSTEM TYPE

Figure 5 — High temperatures (°C and °F) recorded in 10 Rocky Mountain ecosys-
tems. The first graph (A) represents absolute summer highs (ABSMX). The sec-
ond (B) gives average July maxima (AVJMX). The ecosystem types, symbols, and
lines are as in figure 1 .

147

ECOSYSTEM TYPE

Figure 6— Biomass production in 10 Rocky Mountain ecosystems (see also table 2). The
first graph (A) presents harvestable production (HPRD, g/m^/yr). The second (B) gives to-
tal production (TPRD, g/m^/yr). The ecosystem types, symbols, and lines are as in
figure 1 .

Drought likely sets the lower physiological limit
(Hutchinson 1958) for whitebark pine's downward exten-
sion. Walter's (1973) index suggests the drought months
and water deficits are near zero in high forests (whitebark
pine, subalpine fir, and Douglas-fir), greater in ponderosa
pine forests and grasslands, and still greater in Atriplex
shrublands (fig. 4). If (as noted above) wind decreases
and waterholding capacity of soils increases downslope,
water conditions for whitebark production improve as one
moves downslope; this is consistent with the fact that
whitebarks growing in the subalpine fir zone can be stately
timber producers (Pfister and others 1977; Weaver and
Dale 1977). Water deficits remain slight down through the
subalpine fir and possibly the Douglas-fir zones. The com-
plete absence of whitebarks in the ponderosa pine and
grassland zones, despite possible nutcracker dispersal
(Lanner 1990; Tomback and others 1990), supports my
doubt that the tree is capable of growing in the ponderosa
pine and drier grassland-desert zones. Similarly, in the
geographic dimension, the southern and eastern limits of
whitebark pine seem to appear where precipitation in its
altitudinal zone becomes too low (Arno and Weaver 1990).

The lower realized limit (Hutchinson 1958) of white-
bark is set by competition with subalpine fir and lodge-
pole pine rather than drought. In our region, evidence for
this fact lies in the tree's existence and good growth on
sites in the subalpine fir zone (but apparently not in the
Douglas-fir zone) where subalpine fir and lodgepole com-
petition have been removed by clearcutting or fire. The
lower limit of whitebark is, then, set by those factors —
likely drought (thin soils and wind) and low summer tem-
perature — that exclude subalpine fir and lodgepole pine
from higher sites. In the geographic dimension, to the
north and west where rainfall is higher, whitebark is also
unable to form pure stands or disappears (Arno and
Weaver 1990). (European foresters ask why we don't
eliminate competitive "natives" [subalpine fir and lodgepole
pine] from their zone to allow expansion of whitebark pine
and increased production of quality white pine lumber.)

Production

As one moves from dry grasslands to moist forests and
upward to alpine sites one sees production rise and fall
again. Harvestable production — merchantable standing
crop divided by years in its production — rises from dry
grasslands (100 g/mVyr) to moist forests (300 g/mVyr) and
falls through high forests to the alpine (100 g/mVyr, fig. 6,
table 2). Available data suggest that total aboveground
production (total increment ^ro?lyr) also increases fi-om
dry grass and shrublands (50-100 g/mVjr) to warm forests
(over 500 g/mVyr) and falls to the alpine (100 g/mV5T,
fig. 6, table 1). The trend is parallel, but — at least in for-
ests — total production is far higher than harvestable pro-
duction. Three factors may contribute to this difference:
(1) Much of the production, over the life of a stand, is lost
to needle drop, self pruning, and natural thinning; this
material decomposes and is recycled. The total produc-
tion rates reported are deceptively high because they in-
clude nutrients being recycled to the atmosphere and soil;
that is, this production consists of an actual production
component based on nutrient import (weathering and at-
mospheric delivery) and a restructuring component based
on recycling of nutrients from "obsolete" leaves, branches,
and trees. (2) Only 33 to 50 percent of the material extant
at harvest is merchantable (Weaver and Forcella 1977).
(3) Most of the total production data comes from areas
outside our region. If the climates of Douglas-fir, cedar-
hemlock, and ponderosa pine stands from which produc-
tion data are reported are moister or warmer than the
Montana stands in which the harvestable productivity
data were gathered, somewhat higher productivities
might be expected.

I expect productivity to be controlled, like vegetation
distribution, by climate. In the following paragraphs I
will explore the h3rpothesis that production depends on
growing season length and the warmth of that season.

Growing season length might be indexed as the number
of months where daily average temperatures are above

148

Table 3 — Relationship of production (g/m^/yr) to climatic factors

Harvestable production'

66.00

12.8000

p = 0.50

= 0.05

-171.00

70.0000

p = 0.00

= 0.86

34.00

1.7700

GSxT

p = 0.04

/■2 = 0.44

-38.00

14.4000

GSxQ

p = 0.01

= 0.58

Log HP =

0.78

n - n nn

f2 _ n 7fi

1 75

n nnfin

n — n 1 5

f2 — n 97

1.44

0.0530

GSxQ

p = 0.03

r2 = 0.44

tal Droduction^

-880.00

321.0000

p = 0.16

r2 = 0.22

50.00

8.3200

GSxT

p = 0.32

= 0.12

-398.00

77.0000

GSxQ

p = 0.19

r2 = 0.20

Log TP =

0.40

0.4700

p = 0.03

= 0.45

1.96

0.0086

GSxT

p = 0.32

r2 = 0.12

1.47

0.0820

GSxQ

p = 0.22

^ = 0.22

'Harvestable production (HP, lop diameter greater than 10 cm) and the log of harvestable production
(log HP) are regressed against warm season (WS), warm-moist growing season (GS), the product of
warm-moist growing season and average growing season temperature (GSxT), and the product of
wann-moist growing season and Q,o-based growth support units (GSxQ).

n"otal production (TP) and log total production (log TP) are also regressed against GS, GSxT, and
GSxQ.

0 °C. The rationale is that, iinless air temperatiires are
above 0 °C, exposed soils will be frozen (Weaver, in prepa-
ration) and frozen soils will deliver little water and nutri-
ents. Because a regression of production against this in-
dex of growing season is so loose (r^ = 0.05, table 3, fig. 7),
the hypothesis that warm-season length controls produc-
tion is inadequate.

The previous index may fail because, at least at lower
altitudes, growing season is limited in late summer by a
lack of soil water, rather than temperatiire. If so, the
growing season index would be improved as an index of
production by subtracting soil-dry months from total soil-
warm months. As above, I index drought as periods when
precipitation (mmy2) is less than average temperature (C,
after Walter 1973). Regressions of production against
months without either temperature or moisture stress ex-
plain most of the variance in harvestable production (r^ =
0.86, table 3, fig. 7). Season length is a weaker correlate
of total production (r^ = 0.22). I offer two unsatifying
hypotheses for this important difference: (1) Growing sea-
son predicts harvestable production better than total pro-
duction because photosynthates are allocated first to
canopy maintainence (so this component does not vary
much among types) and second to supporting structure; if
so, the relationship between season length and tnmk pro-
duction is clarified by omitting the common leaf-twig pro-
duction. (2) If total production were measured in slightly
more favorable segments of the environmental types than
harvestable production, the iinderstatement of season
length would weaken the relationship.

If all vegetation — grassland to forest to alpine — had a
single temperature response curve, we would expect pro-
duction in "warm moist soil days" to rise exponentially
with increasing temperatiu-e. This expectation is based

on the rise in chemical reaction rates with rising tem-
perature, on the resultant physiological response curves
(Larcher 1975), on degree-day prediction of plant phenol-
ogy (Chang 1960), and on the prediction that degree-day
formulae might be improved by use of temperatiire aver-
ages computed from a nonlinear Q^^ relationship (Larcher
1975). Thus, we expect production to be better correlated
with the product of growing season days x temperature
than with growing season days alone. In fact, whether
we use our average growing season temperature or grow-
ing season Q index, the relationship is poorer for both
harvestable (r^= 0.44) and total (r^= 0.12) production.
Since the basic physiological response is a physiochemical
necessity, I deduce first that, while the vegetation of each
environmental zone does respond to temperature, it is
specifically adapted to temperattires in that zone and, sec-
ond, that temperature adaptation distinct to each zone
eliminates temperature fi'om the predictive equation. I
doubt that temperature would be eliminated if adaptation
were eliminated by using a genetically homogeneous veg-
etation type; for example, if the region were vegetated
with a single crop, production would be best predicted by
the product of growing season length and temperature.
Since competitiveness of a tree undoubtedly depends on
its productivity relative to competitors, this conclusion
supports earlier speculation that growing season tempera-
ture adaptation provides at least one basis for the
existance of different vegetation zones.

While I have argued that production in a zone is best
predicted from the length of its warm moist growing sea-
son, I expect the standing crop of mature accvmaulative
(woody) vegetation (table 2) to be largely determined by
site fertility. Accmnulation — slow or fast — depends on
photosynthesis minus respiration. Synthesis depends on

149

light, warmth, water (open stomates), and nutrients. In
the absence of stand destruction, annual inputs of light,
warmth, water, and elements with atmospheric cycles
(for example, C, H, O, N) should support eternal accumu-
lation. Limited supplies of nutrients with geologic cycles
(for example, P, K, Ca) will halt accumulation (Weaver
1978). It is also argued (Odum 1969) that accvunulation
is halted when respiration equals production. Thus, in
young trees net production is high because the ratio of
photosynthetic mass to respiratory mass is high, in longer
stemmed trees efflcency declines as the respiratory load
increases, and growth ceases when respiration equals
gross production; heartwood's contribution to this rela-
tionship is in proportion to its inertness. While the respi-
ratory factor undoubtedly contributes to observed declines
in production with stand height (and age) (Weaver and
others 1990), I see nutrient supplies as more limiting in
high forests because the trees are far shorter (and there-
fore more efficient) than productive trees of lower alti-
tudes. In contrast, standing crops in less accumulative
(herbaceous) vegetation are determined by production oc-
curring in one growing season, and while it could be de-
termined by supplies of a geologically cycling nutrient,
standing crop is more likely determined by a bulk re-
source like light (unlikely), warmth, water, or an atmo-
spheric nutrient like nitrogen.

CONCLUSIONS

Whitebark pine vegetation is likely excluded from the al-
pine zone by cool growing season temperatiires or droughts
occurring most likely in winter. It seems unlikely that
freezes of late fall, midwinter, or early spring exclude the
tree from higher sites.

The lower physiological limit of whitebark pine is likely
set by drought. Its lower realized limit is likely due to com-
petition with lodgepole pine or subalpine fir. By control of
competition, managers could probably extend whitebark's
range downslope.

The distribution of other dominant plants on the tem-
perature water gradient may be similarly controlled.

Annual production is strongly related (H = 0.86) to the
number of growing season days. The lack of correlation
with temperature suggests that plants of any vegetation
zone are adapted to temperature conditions peculiar to
that zone. Mature standing crops of woody vegetation are
more likely determined by nonatmospheric nutrients than
temperature and precipitation. Potential standing crops
of herbaceous vegetation, on the other hand, are more
likely limited by bulk resources such as warmth, water,
carbon, or nitrogen.

3 6

WM SEASON, mo

20 40

GS X T (moxC 5)

CL
X

r2= 0.85

J L

3 6

GRO SEASON, mo

GS X Q (moxQ)

Figure 7 — The relationship of harvestable production (HPROD, g/m^/yr) to season
length and temperature: (A) production vs. warm season (WM-SEASON), (B) produc-
tion vs. warm-moist season (GRO-SEASON), (C) production vs. warm-moist season
(GS) X average summer temperature above 5 °C (C-5), (D) production vs. warm moist
season (GS) x growth support units (Qg^). The ecosystem types are as in figure 1 .

150

ACKNOWLEDGMENTS

I thank colleagues who helped me locate weather sta-
tions representative of major environmental types of the re-
gion (S. Amo, D. Despain, J. Habeck, R. Rankin, and others
acknowledged in Weaver 1980 and 1990), acquire the data
(J. Caprio), and choose parameters for study (J. Pickett and
J. Brown). The critical review of P. Burgeron, D. Despain,
S. Harvey, W. Moir, and D. Perry is greatly appreciated. I
am also grateful for the support provided by Intermoimtain
Research Station, USDA Forest Service (INT-92720-
RJVA).

REFERENCES

Aplet, G.; Smith, F.; Laven, R. 1989. Stemwood biomass
and production during spruce-fir stand development.
Journal of Ecology. 77: 70-77.

Armand, A. 1992. Sharp and gradual motmtain timber-
lines as a result of species interaction. In: Hansen, A.;
DiCastri, F. 1992. Landscape boundaries. New York:
Springer Verlag: 360-378.

Amo, S. 1984. Timber line — moimtain and arctic forest
frontiers. Seattle: Moimtaineers. 304 p.

Amo, S.; Weaver, T. 1990. Whitebark pine commimity
types and their pattem in the landscape. In: Schmidt,
W.; McDonald, K, comps. Proceedings — symposium on
whitebark pine ecosystems: ecology and management of
a high-moimtain resource. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest
Service, Intermovmtain Research Station: 97-105.

Burke, M.; Gusta, L.; Quamme, H.; Weiser, C.; Li, P.
1976. Freezing injxiry in plants. Annual Review of Plant
Physiology. 27: 507-528.

Chang, J. 1968. Climate and agriculture. Chicago: Aldine.
304 p.

Collins, D.; Weaver, T. 1978. Effects of summer weather
modification (irrigation in Festuca idahoensis-Agro-
pyron spicatum grasslands. Jovu-nal of Range Manage-
ment. 31: 264-269.

Daubenmire, R. 1943. Vegetation zonation in the Rocky
Mountains. Botanical Review. 9: 325-393.

Daubenmire, R. 1956. Climate as a determinant of vegeta-
tion distibution in eastern Washington and northern
Idaho. Ecological Monographs. 26: 131-154.

Daubenmire, R. 1970. Steppe vegetation of Washington.
Bull. 62. Pullman, WA: Washington Agricultural Ex-
periment Station. 131 p.

Daubenmire, R. 1981. Subalpine parks associated with
snow transfer in the mountains of northern Idaho and
eastem Washington. Northwest Science. 55: 124-135.

Daubenmire, R.; Daubenmire, J. 1968. Forest vegetation
of eastern Washington and northern Idaho. Bull. 60.
Pullman, WA: Washington Agricultural Experiment
Station. 104 p.

Forcella, F.; Weaver, T. 1977. Biomass and productivity of
the subalpine Pinus albicaulis/Vaccinium scoparium as-
sociation, MT USA. Vegetatio. 35: 95-105.

Gams, H. 1931. Die klimatische Begrenzimg von Pflanzen
arealen und die Verteilvmg der hygrischen Konti-
nentalitat in den Alpen. Zietschrit der Gesellschafl fur
Erdkunde. 9: 32-68.

Gates, D.; Papian, L. 1971. Atlas of energy budgets of
plant leaves. London and New York: Academic Press.
277 p.

Geiger, R. 1965. The climate near the ground. Cambridge,

MA: Harvard University Press. 611 p.
Hadley, J.; Smith, W. 1987. Influence of knunmholz

microclimate on needle physiology and survival.

Oecologia. 73: 82-90.
Hanley, D. 1976. Tree biomass and production estimates

for three habitat types in northern Idaho. Bull. 14.

Moscow, ID: University of Idaho, College of Forestry.

16 p.

Hunt, W.; Ingham, E.; Coleman, D.; ElHott, E.; Reid, P.
1988. Nitrogen limitation of production and decomposi-
tion in prairie, moimtain meadow, and pine forest.
Ecology. 69: 1009-1016.

Holdridge, L. 1967. Determination of world plant forma-
tions from simple climatic data. Science. 130: 572.

Huschle, G.; Hironaka, M. 1980. Classification and ordi-
nation of serai plant communities. Journal of Range
Management. 33: 179-182.

Hutchinson, G. 1958. Concluding remarks. Cold Springs
Harbor Symposium in Quantiative Biology. 22: 415-427.

Jacobs, J.; Weaver, T. 1990. Effect of temperature and
temperature preconditioning on seedlimg performance
of whitebark pine. In: Schmidt, W.; McDonald, K,
comps. Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-mountain
resource. Gen. Tech. Rep. INT-270. Ogden, UT: U.S.
Department of Agriculture, Forest Service, Intermoun-
tain Research Station: 134-139.

Kuchler, A. 1964. Potential vegetation of the contermi-
nous US. Spec. Publ. 36. New York: American Geo-
graphical Society. 155 p.

Landis, T.; Mogren, E. 1975. Tree strata biomass of sub-
alpine spmce-fir stands in SW Colorado. Forest Science.
21: 9-12.

Lanner, R. 1980. Biology, taxonomy, evolution, and geog-
raphy of the stone pines of the world. In: Schmidt, W.;
McDonald, K., comps. Proceedings — symposium on
whitebark pine ecosystems: ecology and management of
a high-mountain resource. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Research Station: 14-24.

Larcher, W. 1975. Physiological plant ecology. New York:
Springer. 252 p.

Neilson, R. 1986. High resolution climatic analysis and
southwest biogeography. Science. 232: 27-33.

Nielson, R. 1992. Toward a rule based biome model. Land-
scape Ecology. 7: 27-43.

Nobel, P. 1983. Biophysical plant physiology. San
Francisco: Freeman. 608 p.

Penman, H. 1949. The dependence of transpiration on
weather and soil conditions. Journal of Soil Science.
1: 74-89.

Pfister, R.; Kovalchik, B.; Amo, S.; Presby, R. 1977. For-
est habitat types of Montana. Gen. Tech. Rep. INT-34.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Forest and Range Experiment Sta-
tion. 174 p.

151

Rodin, L.; Bazilevich, N. 1967. Production and mineral cy-
cling in terrestrial vegetation. London: Oliver and Boyd.
288 p.

Salisbury, F.; Ross, C. 1992. Plant physiology. Belmont,
CA: Wadsworth. 682 p.

Scott, D.; Billings, W. 1964. Effects of environmental fac-
tors on standing crop and production of an alpine tun-
dra. Ecological Monographs. 34: 243-270.

Stephenson, N. 1990. Climatic control of vegetation distri-
bution: the role of water balance. American Naturalist.
135: 649-670.

Thilenius, J.; Smith, D.; Brown, G. 1974. Effect of 2,4-D
on composition and production of an alpine plant com-
munity in Wyoming. Journal of Range Management.
27: 140-142.

Thornthwaite, C. 1948. An approach toward a rational
classification of climate. Geographical Review. 38:
55-94.

Tomback, D.; Hoffmann, L.; Sund, S. 1990. Coevolution of
whitebark pine and nutcrackers: implications for forest
regeneration. In: Schmidt, W.; McDonald, K., comps.
Proceedings — sjrmposium on whitebark pine ecosys-
tems: ecology and management of a high-mountain re-
source. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. De-
partment of Agriculture, Forest Service, Intermountain
Research Station: 118-129.

Tranquillini, W. 1979. Physiological ecology of the alpine
timberline. Tree existence at high altitude with special
reference to the European Alps. New York: Springer.
137 p.

U.S. FPL. 1974. Wood handbook: wood as an engineering
material. Agric. Handb. 72. Washington, DC: U.S. De-
partment of Agriculture, Forest Service, Forest Prod-
ucts Laboratory, [n.p.].

USGS. 1970. The national atlas. Washington, DC: U.S.
Department of the Interior, Geological Survey. 417 p.

Walter, H. 1973. Vegetation of the earth in relation to cli-
mate and ecophysiological conditions. New York:
Springer. 237 p.

Weaver, T. 1978. Changes in soils along a vegetation-
altitudinal gradient of the Northern Rocky Mountains.
In: Youngberg, C, ed. Forest soils and land use. Pro-
ceedings; 5th North American Soils Conference. Fort
Collins, CO: Colorado State University, Forestry
Department: 14-29.

Weaver, T. 1980. Climates of vegetation types of the
northern Rocky Mountains and adjacent plains. Ameri-
can Midland Naturalist. 103: 392-398.

Weaver, T. 1983. Yield response of high plains grasslands
to water enrichment, three phases. In: Holman, L.;
Knutsen, G. State of Montana activities in the High
Plains Cooperative Program 1981-1983. Helena, MT:
Montana Department of Natural Resources, Water
Resources Division: 77-94.

Weaver, T. 1990. Climates of subalpine pine woodlands.
In: Schmidt, W.; McDonald, K., comps. Proceedings —
S3mtiposium on whitebark pine ecosystems: ecology and
management of a high-mountain resource. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
72-79.

Weaver, T.; Collins, D. 1977. Effects of weather modifica-
tion (increased snow pack) on Festuca idahoensis mead-
ows. Journal of Range Management. 30: 451-456.

Weaver, T.; Dale, D. 1974. Pinus albicaulis in central
Montana: environment, vegetation, and production.
American Midland Naturalist. 92: 222-229.

Weaver, T.; Forcella, F. 1977. Biomass of fifty conifer for-
ests and nutrient exports associated with their harvest.
Great Basin Naturahst. 37: 395-401.

Weaver, T.; Forcella, F.; Dale, D. 1990. Stand develop-
ment in whitebark pine woodlands. In: Schmidt, W.;
McDonald, K., comps. Proceedings — symposium on
whitebark pine ecosystems: ecology and management
of a high-mountain resource. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Research Station: 151-155.

West, N. 1983. Temperate deserts and semideserts: Inter-
mountain salt desert shrubland. Amsterdam: Elsevier:
382.

Whittaker, R. 1975. Communities and ecosystems. New

York: Macmillan. 385 p.
Whittaker, R.; Niering, W. 1975. Vegetation of the Santa

Catalina Mountains, Arizona. V. Biomass, production

and diversity along the elevation gradient. Ecology. 56:

771-790.

Willms, W.; Smoliak, S.; Bailey, A. 1986. Herbage produc-
tion following litter removal on Alberta native grass-
lands. Journal of Range Management. 39: 536-540.

152

Regeneration

153

VARIATION IN SIZE AND WEIGHT OF
CONES AND SEEDS IN FOUR NATURAL
POPULATIONS OF CARPATHIAN
STONE PINE

I. Blada
N. Popescu

Abstract — High variation was found within populations of
stone pine (Pinus cembra L.) for seeds per cone and seeds per
cone weight, but middle high for 1,000-seed weight and low for
cone length and cone diameter. The cone length, cone diameter,
seeds per cone, seeds per cone weight, and 1,000-seed weight
means were 4.7 cm, 4.2 cm, 37.8 cm, 9.2 g, and 250 g, respec-
tively. Southern populations were similar in cone length and
cone diameter, but not in seeds per cone, seeds per cone weight,
and l.OOO-seed weight. Cone length and cone diameter from
open-pollinated, cross-pollinated and self-pollinated cones were
similar, while seeds per cone, seeds per cone weight, and 1,000-
seed weight, were not. All traits displayed continuous variation.
Significant correlations were found among all traits, except
1,000-seed weight.

Stone pine (Pinus cembra L.) is naturally distributed
in the highest forest zone of the Alps and Carpathian
Mountains (Critchfield and Little 1966; Holzer 1975).
In the Alps, the low-elevation stands range between 1,100
to 1,500 m, but the main zone extends between 1,700 to
2,000 m (Contini and Lavarelo 1982; Holzer 1975), while
the high-elevation form of the species climbs as single
trees up to 2,700 m above sea level (Moser 1960). In
Romania, stone pine ranges from 1,350 to 1,880 m in the
northern Carpathians (Gubesch 1971) and from 1,350 to
1,980 m in the southern Carpathians (Beldie 1941; Oarcea
1966; Tataranu and Costea 1952).

Because of its tolerance for low temperatvires, the spe-
cies is very important for reforestation of the subalpine
zone; in this zone it is also important on watersheds, for
stabilizing avalanche areas, and for reducing the effects
of flash floods (Holzer 1972).

Stone pine has a particular importance for the silvicul-
ture of the subalpine zone of the Carpathians. For this
reason a genetic improvement program with both intra-
and interspecific crosses is being carried out (Blada 1982,
1990a); some results have been published (Blada 1987,
1990b, 1992a) or are in preparation (Blada 1992b).

This paper reports on the phenotypic variation in size
and weight of cones and seeds in four natural populations
of Carpathian stone pine, as part of the program.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

I. Blada and N. Popescu are Forest Geneticists, Forest Research Insti-
tute, Bucharest 11, Romania.

MATERIALS AND METHODS

Twenty trees were sampled at random in each of the
four populations listed in table 1. In mid-July 1991, 25
cones from open pollination (OP) on each of the 20 trees
(total 80 trees) were protected against the mountain jay
(Nucifraga caryocatactes L.) by using metal net bags 25
by 20 cm in size.

For comparative reasons, cones and seeds obtained from
controlled cross- and self-pollinations (CP, SP) in the
Gemenele population were measured; these pollinations
were performed in July 1989 using 10 x 10 full diallel
mating design (Blada 1992b), according to Griffing's
(1956) Experimental Method 1.

Cone and seed measurements were taken soon afler
their collection in October 1990 for controlled cross-
pollinated and self-pollinated and in October 1991 for
open-pollinated cones. Twenty-five cones were measured
from each tree resulting fi*om open-pollinated cones, and
20 cones fi-om each combination resulting fi*om controlled
cross-pollinated and self-pollinated cones. Five traits
were measured as shown in table 2.

Using data from measurements, the population mean
(x), standard deviation (g), mean square {&), variation co-
efficient (VC), range of variation (Q), and correlation coef-
ficient (r) were calculated.

WITHES-POPULATION VARIATION

Statistical parameters of the cones and seeds of the four
Carpathian stone pine populations listed in table 1 are
summarized in table 3 and figure 1.

Table 1 — Geographic distribution of Pinus cembra studied
populations in the Carpathians

Population

Latitude N. Longitude E.

Altitude

Carpathian
Range

Degrees

Meters

Gemenele

45°35" 22°50'

1,720

Southern

Stana de Rau

45°25' 23°03'

1,450

Southern

Pietrele

45°23' 22°52'

1,550

Southern

Lala

47°33' 25°05'

1,520

Northern

154

Table 2 — Measured traits

Trait Units Symbols

Cone length Centimeters CL

Cone diameter Centimeters CD

Seeds per cone Number SC

Seeds per cone weight Grams SCW

1 ,000-seed weight Grams 1 ,000-SW

The following main results were obtained for the
Gemenele, Stana de Rau, Pietrele, and Lala populations
(in sequence):

• The cone length mean was 4.9 ± 0.4 cm, 4.7 ± 0.4 cm,
4.9 ± 0.6 cm, and 4.1 ± 0.4 cm, respectively.

• The cone diameter mean was 4.2 ± 0.3 cm, 4.1 ± 0.3 cm,
4.1 ± 0.3 cm, and 3.5 ± 0.2 cm, respectively.

• The seeds per cone mean was 40.9 ± 9.6, 36.7 ± 8.1,
52.6 ± 13.3, and 21.1 ± 7.1, respectively.

• The seeds per cone weight was 9.8 ± 2.7 g, 9.0 ± 2.5 g,
12.5 ± 3.9 g, and 5.6 ± 1.7 g, respectively.

Table 3 — Mean (x), phenotypic standard deviation (a), mean square
(cf), variation coefficient {VQ, range of variation (O)

Parameters (open pollination)

Trait

x± a

Gemenele population

4.9

± 0.4

0.15

7.9

4.1

5.8

4.2

+ 0.3

0.11

7.9

3.6

5.0

40.9

± 9.6

91.31

23.3

22.5

60.2

SCW

9.8

± 2.7

7.50

28.0

4.6

18.1

1,000-SW

238.0

± 47.4

2,243

19.9

170

352

Stana de Rau population

4.7

± 0.4

0.18

9.1

3.8

5.6

4.1

± 0.3

0.08

7.0

3.7

4.9

36.7

± 8.1

65.67

22.1

21.7

55.4

SCW

9.0

± 2.5

6.05

27.3

5.5

14.1

1,000-SW

252

± 37.6

1,412

14.9

175

293

Pietrele population

4.9

± 0.6

0.34

11.7

4.1

6.0

4.1

± 0.3

0.10

7.7

3.4

4.7

52.6

± 13.3

177.31

25.3

31.2

90.4

SCW

12.5

± 3.9

15.27

31.3

6.2

23.1

1 ,000-SW

238

± 60.9

3,712

25.6

138

428

Lala population

4.1

± 0.4

0.17

9.9

3.7

5.1

3.5

± 0.2

0.04

6.0

3.1

3.9

21.1

± 7.1

51.15

34.0

13.6

34.8

SCW

5.6

± 1.7

2.78

29.9

3.6

9.4

1 ,000-SW

270

± 46.5

2,163

17.2

184

346

Trait parameters across four populations

4.7

± 0.5

0.30

11.7

3.7

6.0

4.2

± 0.3

0.11

7.9

3.1

5.0

37.8

± 14.9

222.22

39.4

13.6

90.4

SCW

9.2

± 3.7

13.77

40.3

3.6

23.1

1 ,000-SW

250

± 49.7

2,471.00

19.8

167.0

354.0

• The 1,000-seed weight mean was 238 + 47.4 g, 252 ±
37.6 g, 238 ± 60.9 g, and 270 ± 46.5 g, respectively.

Therefore, the southern populations were similar in
cone length and cone diameter, but not in seeds per cone,
seeds per cone weight, and 1,000-seed weight. On the
other hand, the southern populations ranked high and the
northern population ranked low for all measured traits.

Very high variation coefficients were found within
each population for both seeds per cone and seeds per
cone weight, but middle high for 1,000-seed weight and
low for cone length and cone diameter traits.

The last three columns of table 3 give information con-
cerning mean squares, variation coefficients, and the
range of variation for pop\ilation traits.

ACROSS-POPULATION
PARAMETERS

The mean values and ranges of the traits across foiu*
studied populations — and by extrapolation — for all the
Carpathian stone pine were as follows (table 3, lower
part):

• The cone length mean ranged from 3.7 to 6.0 cm, with
a mean of 4.7 ± 0.5 cm, and the cone diameter ranged
from 3.1 to 5.0 cm, with a mean of 4.2 ± 0.3 cm. According
to Contini and Lavarelo (1992), the size of cones from the
Alps ranged from 5.0 to 10.0 cm in length and from 4.0 to
6.0 cm in diameter. Therefore, the cone size from the Alps
ranked high and the cones from the Carpathians ranked
low.

• The seeds per cone ranged from 13.6 to 90.4, with a
mean of 37.8 ± 14.9. However, the seeds per cone mean
from the Alps varied between 46 and 164 with a mean
of 93 seeds (Rohmeder and Rohmeder 1955). Thus, the
seeds per cone from the Alps ranked high and the seeds
per cone from the Carpathians ranked low.

• The seeds per cone weight mean ranged from 3.6 to
23.1 g, with a mean of 9.2 ± 3.7 g.

• The 1,000-seed weight mean ranged from 167 to
354 g, with a mean of 250 ± 49.7 g. Consequently, a kilo-
gram of seed from the Carpathians could include from
2,825 to 5,988 seeds, with a mean of 4,000 seeds. Accord-
ing to Rohmeder and Rohmeder (1955), the 1,000-seed
weight mean from the Alps ranged between 150 and
350 g; thus, the Carpathians 1,000-seed weight mean
was very close to that of the Alps.

The cone length, cone diameter, seeds per cone, seeds
per cone weight, and 1,000-seed weight coefficients ac-
counted for 11.7 percent, 7.9 percent, 39.4 percent, 40.3
percent, and 19.8 percent of the variation, respectively
(table 3, lower part).

POLLINATION COMPARISONS

The cone and seed parameters shown in table 4 came
from records of 2,000 cones from 80 open-pollinated (OP)
trees, 1,800 cones from a 10 x 10 full diallel mating design
for cross-pollination (CP), and 200 cones from 10 self-
pollinated (SP) trees, all from the Gemenele population.
For example, the cone length ranged from:

155

300,

250.

200.

.C150.

Jioo

o
o

° 50

;

s
s
s
s

1 2 3 4 X X X

1 2 3 A X X X

2 3 A X X X

1 2 3 A 5f X f

Figure 1 — Mean performance of five measured traits of cones and seeds from open pollination_
in four populations (1 , 2, 3, 4) compared to controlled cross-pollination (f ) and self-pollination (^)
(x= average across 1 , 2, 3, 4 populations).

4.1 to 5.8 cm, with a mean of 4.9 ± 0.4 cm for open-
pollinated;

3.5 to 6.0 cm, with a mean of 4.8 ± 0.6 cm for cross-
pollinated;

3.7 to 5.5 cm, with a mean of 4.7 ± 0.6 cm for self-
pollinated.

Variation coefficients (VC) for open-pollinated, cross-
pollinated, and self-pollinated were 7.9 percent, 12.8 per-
cent, and 12.1 percent, respectively.

Table 4 — Comparison between statistical parameters of cones and seeds in
the Gemenele population, according to the pollination type

Type of Parameters

Trait

pollination^

k±a

4.9

0.4

0.15

7.9

4.1

5.8

4.8

0.6

0.37

12.8

3.5

6.0

4.7

0.6

0.36

12.1

3.7

5.5

4.2

0.3

0.11

7.9

3.6

5.0

4.3

0.4

0.15

9.1

3.7

5.4

4.2

0.4

0.16

9.4

3.6

4.9

40.9

9.6

91.31

23.3

22.5

60.2

62.0

16.5

272.24

26.6

41.4

96.4

64.0

12.9

167.12

20.3

44.9

85.7

sew

9.8

2.7

7.50

28.0

4.6

18.1

14.1

4.4

19.56

31.3

7.0

24.5

13.8

3.1

9.66

22.5

9.5

17.6

1,000-SW OP

238.0

47.4

2,243.00

19.9

170.0

352.0

188.0

31.3

978.00

16.6

11.7

26.5

179.0

25.6

656.94

14.3

14.7

22.3

'OP = open pollination; CP = controlled cross-pollination in a 10 x 10 diallel; SP = self
pollination of 10 parents.

Similar data can be foimd for cone diameter, seeds per
cone, seeds per cone weight, and 1,000-seed weight in
table 4.

CORRELATIONS

Significant (p < 0.05) and highly significant (p < 0.01)
correlations were found between cone length and cone di-
ameter, cone length and seeds per cone, cone length and
seeds per cone weight; cone diameter and seeds per cone,
cone diameter and seeds per cone weight; seeds per cone
and seeds per cone weight. No significant correlation
was found between 1,000-seed weight and any other trait
(table 5). These strong correlations among the main cone
and seed traits suggest that improvement (quantitatively)
of seed production could be made by indirect selection; for
example, selection for cone length, as an easily measur-
able trait, will cause an increase in seeds per cone and
seeds per cone weight, and consequently in seed production.

TYPE OF DISTRIBUTION

The distribution frequencies of cone length, cone diam-
eter, seeds per cone, seeds per cone weight, and 1,000-
seed weight were very close to normal distribution (fig. 2).
According to genetic theory (Mather and Jinks 1977), this
pattern of distribution is specific to quantitative traits.
Such traits are polygenically controlled.

156

Table 5 — Phenotypic correlations among Pinus cembra cone and
seed traits (Df = 1 8)

Trait

sew

1 nnn.sw

1.000

0.577-

0.636-

0.669-

0.272

1.000

0.499*

0.647-

0.227

1.000

0.636-

0.325

sew

1.000

0.396

1,000-SW

1.000

DISCUSSION

Our observations in the main Carpathian populations
indicated that tree, locahty, and year in which the cones
were initiated may significantly affect cone and seed
traits. As the measurements were performed on cones
collected in two different years, the comparisons were
perhaps not entirely valid.

The seeds per cone and seeds per cone weight from open-
pollinated cones ranked lower than the same parameters
from both cross-pollinated and self-pollinated cones; but,
surprisingly, 1,000-seed weight from open-pollinated
cones ranked higher than 1,000-seed weight from cross-
pollinated cones. The lack of significant correlation
between 1,000-seed weight and seeds per cone weight
(table 5) partially explains this unexpected resvilt. Also,
these differences could be attributable to the biological
and climatic factors (temperature, moistm-e, and wind)
that occurred in 1990 and 1991.

The cone length and cone diameter means from self-
pollinated cones were similar to cone length and cone
diameter means from both open-pollinated and cross-
pollinated cones, but the seeds per cone and seeds per
cone weight means from self-pollinated cones were similar
to seeds per cone and seeds per cone weight means from
cross-pollinated cones and ranked higher them open-
pollinated cones. Consequently, stone pine was found

£ 20

^ 20 -

£ 10

330

3.7 3.9 4.2 4.5 4.8 5.1 5.4 5.7 6.0

Cone Length (cm)

3.1 3.4 3.7 4.0 4.3 4.6 4.9

Cone Diameter (cm)

^ 20

S> 10

10 20 30 40 50 60 70 80 90

Seeds per Cone (No.)

30 -

^ 20 ■

2> 10

4 7 10 13 16 19 22

>g 20-

150 190 230 270 310 350 390 430

Seeds per Cone Weight (g) 1 ,000-Seed Weight (g)

Figure 2— Frequency distribution in the Pinus cembra populations evaluated for five traits.

157

to be highly self-fertile. This led to the conclusion that
a very high proportion of the seeds produced by wind pol-
lination would be selfs; this is an undesirable characteris-
tic since selfed seed produces slower growing trees and
the seedlings have a lower survival rate.

Finally, it should be stressed that although the stone
pine cone diameter, cone length, and seeds per cone
means from the Alps exceeded the same trait means from
the Carpathians, the 1,000-cone weight mean from the
Carpathians was similar to that from the Alps.

CONCLUSIONS

Southern populations were similar in cone length and
cone diameter but not in seeds per cone, seeds per cone
weight, and 1,000-seed weight. Southern populations
ranked high and the northern ones ranked low in all
measured traits.

Very high variation in both seeds per cone and seeds
per cone weight was found within each population, but
variation in 1,000-seed weight was moderate, and it was
low in cone length and cone diameter traits.

Cone length and cone diameter from open-pollinated,
cross-pollinated, and self-pollinated cones were similar,
while seeds per cone, seeds per cone weight, and 1,000-
seed weight were not.

The strong correlations among the main cone and seed
traits suggest that genetic improvement in seed produc-
tion could be attained by indirect selection.

All measured traits displayed a continuous variation,
suggesting polygenic control.

ACKNOWLEDGMENTS

Gratitude is expressed to Dr. H. Kriebel from Ohio
State University for the revision of the English version
of the paper and for the useful suggestions.

REFERENCES

Beldie, A. 1941. Observatii asupra vegetatiei lemnoase
din Muntii Bucegi. Analele ICEF, Seria 1, vol. 6: 39-43.

Blada, I. 1982. Relative blister rust resistance of native
and introduced white pines in Romania. In: Heybroek, H.;
[and others], eds. Resistance to diseases and pests in
forest trees. PUDOC, Wageningen, The Netherlands:
415-416.

Blada, I. 1987. Genetic resistance to Cronartium ribicola
and height growth in some five needle pines and of

some inter-specific hybrids. Bucharest: Academy of
Agricultural and Forestry Sciences. 146 p. Thesis.

Blada, I. 1990a. Breeding Pinus cemhra by intra- and
inter-specific crosses. For. Res. Inst., Bucharest, Re-
search Plan, No. 7/1990.

Blada, I. 1990b. Genetic variability of some traits in two
Pinus cembra natural populations. In: Garrett, P. W.,
ed. Proceedings of a s3miposium on white pines prov-
enances and breeding. Gen. Tech. Rep. NE-155. Radnor,
PA: U.S. Department of Agriculture, Forest Service,
Northeastern Forest Experiment Station: 56-68.

Blada, I. 1992a. Pinus cembra and its inter-specific hy-
brids. Paper presented at the lUFRO Centennial Meet-
ing, Berlin-Eberswalde, August 31-September 5, 1992.

Blada, I. 1992b. Diallel cross in Pinus cembra. I. Variation
in cone and seed traits. [In preparation].

Contini, L.; Lavarelo, Y. 1982. Le pin cembro: repartition,
ecologie, sylviculture et production. Paris: INRA, ISBN.
197 p.

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis-
tribution of the pines of the world. IVIisc. Publ. 991.
Washington, DC: U.S. Department of Agriculture, For-
est Service. 97 p.

Griffing, B. 1956. Concept of general and specific com-
bining ability in relation to diallel crossing systems.
Australian J. Sci. 9: 465-493.

Gubesch, L. 1971. Raspandirea relictului glaciar zambrul
(Pinus cembra) pe versantii sudici ai unor masive din
Calimani. Ocrotirea Naturii, 15, 2: 149-159.

Holzer, K. 1972. Intrinsic qualities and growth potential
of Pinus cembra and P. peuce in Europe. In: Bingham,
R. T.; [and others], eds. Biology of rust resistance in for-
est trees. Misc. Publ. Washington, DC: U.S. Department
of Agriculture, Forest Service: 99-110.

Holzer, K. 1975. Genetics of Pinus cembra L. Anales
Forestales, Zagreb. 6/5: 139-158.

Mather, K.; Jinks, J. L. 1977. Introduction to biometrical
genetics. Chapman and Hall, Ltd. 233 p.

Moser, L. 1960. Verbreitung und Bedentung der Zirbe
im italienishcen Alpengebiet. Jb. Ver. Z. Schutz
Alpenpfl.u.-tiere. 25: 16-21.

Oarcea, Z. 1966. Contributii la cunoasterea raspandirii si
vegetarii pinului cembra in Retezat. Revista Pad. 9:
495-497.

Rohmeder, E.; Rohmeder, M. 1955. Untersuchungen uber
das Samentragen und Keimen der Zirbelkiefer (Pinus
cembra) in den Bayrischen Alpen. Allg. Forstzeitschrift.
10: 83.

Tataranu, I. D.; Costea C. 1952. Un arbore de interes
forestier: Pinus cembra. Revista Padurilor. 11: 3-14.

158

ROLE OF NUTCRACKERS ON SEED
DISPERSAL AND ESTABLISHMENT OF
PINUS PUMILA AND P. PENTAPHYLLA

Mitsuhiro Hayashida

AbBtract— Nutcrackers (Nucifraga caryocatactes) harvest almost
all Pinus pumila Kegel cones and cache the seeds in the soil. Most
P. pentaphylla 'M.ayr seeds scatter after the cones open, while nut-
crackers collect the seeds still held in the cones and then cache the
seeds. Many seedlings of P. pumila and P. pentaphylla are often
found in clusters consisting of several individuals of the same age.
Almost all P. pumila seedlings and clusters of P. pentaphylla
seedlings on barren slopes likely originate from seeds cached by
nutcrackers.

Stone pines (subsection Cembrae in Pinus) have large
wingless seeds that are dispersed by caching of vertebrates
(Hayashida 1989a; Hutchins and Lanner 1982; Mattes
1982; Tomback 1982). Nutcrackers {Nucifraga spp.) are
the main dispersal agents of seeds of these pines (Lanner
1989). The nutcracker also caches other large wingless
seeds or short-winged seeds in section Strobus (Lanner
and Vander Wall 1980; Tomback 1990; Vander Wall and
Balda 1977).

Pinus pumila Regel (Japanese stone pine) is a dwarf pine.
Its range extends through Japan and Korea into Siberia,
Kamchatka (Critchfield and Little 1966). On most of the
high mountains in Japan, a vegetational zone dominated
by P. pumila occurs above the forest limit. This area is
called the "Pinus pumila zone" by Japanese ecologists. The
regeneration of P. pumila thickets is mainly due to a veg-
etative regeneration by adventitious roots (Okitsu and Ito
1983). Large wingless seeds of P. pumila are known to be
dispersed by nutcrackers (Saito 1983; Turcek and Kelso
1964), but a detailed observation for seed dispersal and
caching behavior is needed.

Pinus parviflora Sieb. et Zucc. (Japanese white pine) is
distributed throughout the Japanese islands (Critchfield
and Little 1966). This species consists of two geographical
varieties, which intergrade in central Japan (Hayashi 1954).
The northern variety is sometimes called P. pentaphylla
Mayr; the southern P. himekomatsu Miyabe and Kudo or
a variety of the former. Since the seed wings of P. penta-
phylla are longer than those of P. himekomatsu (Ishii 1968),
I distinguished between the two pines. There are few eco-
logical studies of these pines except for community struc-
ture (Tatewaki and others 1960; Yoshioka and Saito 1962).

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Mitsuhiro Hayashida is Associate Professor, Department of Bioenviron-
ment, Faculty of Agriculture, Yamagata University, Wakaba, Tsuruoka
997, Japan.

Pinus pumila and P. pentaphylla are in the subgenus
Strobus and are so similar that they naturally hybridize
(P. hakkodensis Makino) (Ishii 1968); but these pines differ
in seed, cone, and other characteristics. The objectives of
this paper are to describe the processes from seed dispersal
to seedling estabUshment of P. pumila and P. pentaphylla
and to examine the role of nutcrackers in their regeneration.

STUDY AREA

The study was conducted on Mount Apoi (811 m above sea
level), which is located in south-central Hokkaido, Japan
(42°6' N., 143°2' E.). It is composed of ultramafic rocks
(dunite, IherzoHte, etc.) (Niida 1984), and is covered with
coniferous forests. The P. pumila zone extends from eleva-
tions of 500 m to 800 m. Below 500 m, P. pentaphylla for-
est occupies the rocky slopes and ridges, while the greater
part of the mountain is covered by coniferous forests domi-
nated by Picea glehnii, Pinus pentaphylla, and Abies
sachalinensis (Hayashida 1989b).

METHODS

The number and distribution of 2-year cones were mapped
to determine rates of seed predation by vertebrates and in-
sects. These cones were counted from July to October in
1984-87. The observations on seed harvesting, transport-
ing, caching, and retrieving behavior in vertebrates were
made in 1984-87 using binoculars and a telescope.

Many seedlings of P. pumila and P. pentaphylla were
found on barren slopes along mountain trails. These seed-
lings were often found in clusters, which consisted of sev-
eral seedhngs of the same age. To obtain the frequency
distribution of number of individuals per cluster in both
pines, all seedlings on barren slopes along mountain trails
were recorded.

SEED AND CONE MORPHOLOGY

Pinus pumila has wingless seeds; P. pentaphylla seeds
have wings that are seed length (table 1). Seeds of both
pines have the same thick coats. Pinus pumila seeds are
significantly heavier than those of P. pentaphylla, though
the seeds of P. pentaphylla are larger.

Pinus pumila cones do not open when they ripen, but
P. pentaphylla cones open at maturity and release seeds
to fall free. Pinus pumila cones weigh less than P. penta-
phylla cones, but contain more seeds.

159

Table 1 — Seed and cone characteristics of Pinuspumila and P. pentaphylla

Pinus pumila P. pentaphylla

MAAQIirOITIAnt

d 1 Id It

Moan + Qn

Moon 4- QD

Seed length (mm)

7.700 ± 0.600

105

10.000 ± 0.700

105

0.001

Dry weight (g)

0.100 ± 0.014

0.068 ± 0.014

0.001

Seed coat thickness (mm)

0.390 ± 0.040

Length of wings (mm)

wingless

10.500 + 1.600

105

Cone mass (g)

11.00013.300

24.800 ± 7.000

0.001

Number of seeds per cone

42.800 ±12.900

28.900 ±9.100

0.001

Dehiscence

not open

open

'Mann-Whitney U test.

SEED DISPERSAL OF P. PUMILA

Figure 1 shows the seasonal disappearance of 2-year
cones of P. pumila in 1985-88. All cones disappeared from
trees by mid-October every year. I observed nutcrackers
(Nucifraga caryocatactes), squirrels (Sciurus vulgaris ori-
entis), and chipmunks (Tamias sibiricus lineatus) harvest-
ing cones. Five species of vertebrates harvested and car-
ried seeds to caches: nutcracker, vsiried tit (Parus varius),
nuthatch (Sitta europaea), squirrel, and chipmunk. The
nutcracker was the most frequently observed vertebrate
carr3dng P. pumila seeds.

Nutcrackers usually harvest seeds by removing the
cones and flying with them to perches where they extract
the seeds and hold them in their sublingual pouches. Var-
ied tits and nuthatches were able to take P. pumila seeds
exposed by nutcracker foraging and eat them. They were
seen caching pine seeds on a tree or in the soil. Squirrels
visited the P. pumila scrub from a coniferous forest to for-
age on pine seeds. They often carried a whole cone to cache
it in scatter-hoarding. Chipmunks harvest pine seeds in
their cheek pouches. They carried about 50 seeds on a trip
and cached them in the soil.

These results indicate that nutcracker is the main seed
disperser of P. pumila.

E
«
oc

Jul.

Aug.

Sep.

Oct.

Figure 1 — Retention of mature cones of Pinus pumila
on trees (in years 1985-88) as related to season.

SEED DISPERSAL OF
P. PENTAPHYLLA

Cones taken by vertebrates or damsiged by insects
(cone-boring lepidopteran larvae) accounted for less than
20 percent of the total. Usually, nearly 90 percent of the
cones open on the trees from mid-September through early
October. After opening, about 80 percent of the seeds im-
mediately scatter to the ground, due to the fact that 77
percent of P. pentaphylla cones were oriented downward
(Hayashida 1989b).

Nutcrackers, varied tits, and nuthatches harvested and
cached seeds still held in the upward-oriented open cones.
In the years when P. pumila crop sizes were small, nut-
crackers harvested P. pentaphylla seeds more frequently.
They extracted the several seeds from an open cone, peeled
off the seed wings, and put seeds into their sublingual
pouches. They moved rapidly around a tree and went from
tree to tree.

NUTCRACKER CACHING BEHAVIOR

Nutcrackers began to cache P. pumila seeds in early
August when the seeds were mature, and they continued
to harvest seeds until the seed crop was depleted. They
cached the seeds in the soil at a depth of 2 to 3 cm. Cache
size (the number of seeds per cache) ranged from 1 to 51.
Nutcrackers cached P. pumila seeds mainly in coniferous
forests.

In mid-September, P. pentaphylla cones began to open,
then nutcrackers began to harvest seeds from the opening
cones. Nutcrackers scatter-hoarded almost all seeds in the
soil (97 percent) except for two caches containing one and
two seeds that were established in the bark on fallen trees.
Cache size ranged from 1 to 40 seeds. Seventy-eight caches
were recorded in coniferous forests and only one cache in
the Pinus pumila zone.

CLUMPS OF SEEDLINGS

Eighty-four percent of the occurrence of P. pumila seed-
lings were in groups of two or more trees (table 2). These
seedlings were 1 km in distance from the nearest seed tree.

160

Table 2 — Number of individuals per clump of Pinus pumila and P. pentaphylla on
barren slopes along mountain trails

Number of clumoQ

Numbor of inHiviclijalQ

(percent)

per clump

Species

Single Clump (2<)

Maximum Mean ± SD

Pinus pumila

9(16) 47(84)

26 5.6 ±5.3

P. pentaphylla

137(72) 54(28)

25 2.6 ± 3.8

Twenty-eight percent of P. pentaphylla seedlings were
growing in groups of two or more trees. There were many
seed trees of P. pentaphylla along the moimtain trails. Sin-
gle seedlings might originate from naturally scattered fallen
seeds, and the seedlings in cliunps probably originated from
caches by nutcrackers.

DISCUSSION

All mature cones of P. pumila are harvested by verte-
brates every year (fig. 1). My observations indicate that
almost all of the transported seeds were carried by nut-
crackers. Nutcrackers cached seeds in the soil at a depth
of 2 to 3 cm. It is shallow enough to permit seedling estab-
lishment. Therefore, seed dispersal of P. pumila is heavily
dependent on seed caching by nutcrackers.

Nutcrackers carried and cached P. pumila seeds mainly
into coniferous forests. In general, P. pumila seedhngs can-
not grow and establish in coniferous forests even if cached
seeds survive and germinate. However, several P. pumila
trees were found that developed and matured on the rocky
ridges or barren slopes in coniferous forests. These facts
may indicate that the nutcracker is not an efficient seed
disperser of P. pumila under the present conditions. Pinus
pumila adapted its specialized tree form and other features
to the habitat that is characterized by strong winds and
heavy snow accumulation in winter. Thus, the P. pumila
zone is developed on deforested areas in the boreal subal-
pine belt. In Europe, nutcrackers cached P. cembra seeds
in coniferous stands and sometimes cached them above
timberline (Mattes 1982). Seed dispersal over wide areas
by nutcrackers probably played an important role in ex-
panding the range of P. pumila.

Pinus pentaphylla has dehiscent cones and winged seeds.
Most of the seeds scattered when cones opened. Nutcrack-
ers often harvested and cached seeds still held in the up-
ward open cones. About 90 percent of P. pentaphylla trees
found in coniferous forests were individual single trees, but
on the ecotone between coniferous forest and the Pinus
pumila zone, half of the occurrences of trees were in clumps
of two or more individuals (Hayashida 1989b). About 30
percent of the seedlings were growing in clumps on barren
slopes in the coniferous forest zone. These seedlings in
climips probably originated from seeds cached by nutcrack-
ers. Seed dispersal by nutcrackers enables P. pentaphylla
to invade areas that are inaccessible to other coniferous
trees.

The northern variety, P. pentaphylla, has longer seed
wings than the southern variety, P. himekomatsu (Ishii
1968). It is possible that these two varieties have different

morphological features and seed-dispersal syndromes. If
so, why? I would study these problems with an ecological
approach.

ACKNOWLEDGMENTS

I thank Urakawa Local Forest Management Office,
Hokkaido, and Oji Forestry and Landscaping Co. for per-
mission to work in the area and for information on the
area.

REFERENCES

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic distri-
bution of pine of the world. Misc. Publ. 991. Washington,
DC: U.S. Department of Agriculture, Forest Service. 97 p.

Hayashi, Y. 1954. The natural distribution of important
trees, indigenous to Japan. Conifers Rep. 3. Bulletin of
the Government Forestry Experiment Station. 75: 1-173.
[In Japanese with English svmimary].

Hayashida, M. 1989a. Seed dispersal by red squirrels and
subsequent establishment of Korean pine. Forest Ecology
and Management. 28: 115-129.

Hayashida, M. 1989b. Seed dispersal and regeneration
pattern of Pinus parviflora var. pentaphylla on Mt. Apoi
in Hokkaido. Research Bulletins of the College Experi-
ment Forests, Facvdty of Agriculture, Hokkaido Univer-
sity. 46(1): 177-190. [In Japanese with English stmimary].

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Ishii, S. 1968. The basic forestry study on genus Pinus,
especially its taxonomic and geographical consideration.
Journal of the Faculty of Agriculture, Kochi University.
19: 1-114. [In Japanese with English svunmary].

Lanner, R. M. 1989. Biology, taxonomy, evolution, and
geography of stone pines of the world. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposiimi on whitebark pine ecosystems: ecology and
management of a high-mouintain resource; 1989 March
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden,
UT: U.S. Department of Agriculture, Forest Service, In-
termountain Research Station: 14-24.

Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of lim-
ber pine seed by Clark's nutcracker. Journal of Forestry.
78: 637-639.

Mattes, H. 1982. Die Lebensbemeinschaft von Tannenhaher,
Nucifraga caryocatactes (L.), und Arve, Pinus cembra L.,
und ihre forstliche Bedeutiing in der oberen Gebirgs-
waltstufe, Berichte. 241. 74 p.

161

Niida, K. 1984. Petrology of the Horoman utramafic rocks
in the Hidaka Metamorphic Belt, Hokkaido, Japan. Jour-
nal of the Faculty of Science, Hokkaido University, Ser.
IV. 21: 197-250.

Okitsu, S.; Ito, K. 1984. Vegetation d3niamics of the Siberian
dwarf pine (Pinus pumila Kegel) in the Taisetsu moun-
tain range Hokkaido, Japan. Vegetatio. 58: 105-113.

Saito, S. 1983. Caching of Japanese stone pine seeds by
nutcrackers at the Shiretoko Peninsula, Hokkaido. Tori.
32: 13-20. [In Japanese with English summary].

Tatewaki, M.; Tsujii, T.; Kawano, S. 1960. The community
structure and distribution of Pinus parvifLora forest in
Hokkaido. Japanese Journal of Ecology. 10: 120-123. [In
Japanese with English summary].

Tomback, D. F. 1982. Dispersal of whitebark pine seeds by
Clark's nutcracker: a mutualism hypothesis. Journal of
Animal Ecology. 51: 451-467.

Tomback, D. F. 1990. The evolution of bird-dispersed
pines. Evolutionary Ecology. 4: 185-219.

Turcek, F. J.; Kelso, L. 1968. Ecological aspects of food
transportation and storage in the Corvidae. Communi-
cations in Behavioral Biography, Al: 277-297.

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of the
Clark's nutcracker and the pinon pine for efficient seed
harvest £md dispersal. Ecological Monographs. 47: 89-111.

Yoshioka, K; Saito, K. 1962. Differences in the distribution
of forest communities due to topography of the habitats
in Okunikkawa, Miyagi Prefecture. Ecological Review.
15: 213-220.

162

ROLE OF VARIOUS ANIMALS IN
DISPERSAL AND ESTABLISHMENT
OF WHITEBARK PINE IN THE ROCKY
MOUNTAINS, U.S.A.

Harry E. Hutchins

Abstract — The dispersal and establishment of whitebark pine
has been attributed to Clark's nutcracker. But what is the role
of other animals who forage on whitebark pine seed? This study
looks at the role of the diurnal birds and mammals that forage
on whitebark pine cones and seed in western Wyoming. A total
of 1,005 cones were carefully observed in both contiguous "forest"
stands and open-grown "meadow" trees in early November. Red
squirrels harvested most of the whitebark pine seed in forested
plots while Clark's nutcracker removed most of the seed from the
meadow trees. Only a small percentage of seed went to ground
squirrels, chipmunks, Stellar's jays, ravens, etc. The foraging
behavior as well as the importance of these animal species to
whitebark pine regeneration is discussed.

Many animal species have been observed foraging on
the nutritious seeds of whitebark pine (Pinus albicaulis).
But exactly which of these animals plays a role in seed-
ling establishment and to what degree? Which animals
are seed predators? Can whitebark pine seed germinate
from that seed which is dropped through animal foraging
accidents? Do animals harvest all the seed in the fall or
can cones fall to the ground and seed germinate as the
cone decays? This paper will address these questions.

The role of various animals in the dispersal and estab-
lishment of whitebark pine was examined by observing
1,005 cones and determining where and how the seeds
were dispersed. By primarily observing the behavior of
diurnal birds and mammals in the subalpine canopy, I
was able to determine the degree of influence each species
has on seed dispersal and seedling establishment.

METHODS

The study was primarily conducted at the Squaw Basin-
Togwotee Pass area (Bridger-Teton National Forest; fig. 1).
This area of high-elevation meadows offered distant views
of bird activity, whitebark pine growing in both contigu-
ous forest and open-grown situations, and an abundant
cone crop for whitebark pine diuing 1980. Besides white-
bark pine, the forest stands were composed of Engelmann
spruce (Picea engelmannii), subalpine fir {Abies lasio-
carpa), and a minor amoimt of lodgepole pine (Pinus con-
torta). The moraine meadow ridges were pioneered by

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Harry E. Hutchins is Natural Resource Instructor, Itasca Community
College, Grand Rapids, MN.

lone whitebark pine. This provided an opportimity to
study two different types of commimities — "forest" and
"meadow" (fig. 2).

Observations of animal activity were also made at
Mount Washburn (2,680 to 3,140 m, Yellowstone National
Park, WY) and Surprise Lake (2,960 m. Grand Teton Na-
tional Park, WY). These areas were similar to the Squaw
Basin site except they lacked open-grown, cone-bearing
trees. Details on the study sites can be foxmd in Hutchins
and Lanner (1982) and Vander Wall and Hutchins (1983).

Several cone-bearing whitebark pines were chosen for
observation in both continuous forest in which red squir-
rels {Tamiasciurus hudsonicus) were present and in the
open meadow, at least several hundred meters from the
nearest forest edge (these sites lacked squirrels). In each
of these two types of sites, 1,005 whitebark pine cones
were observed during the period Jvdy 3 to November 2,
1980. The cones were scattered among several mature
trees and were counted at 1- to 2-week intervals using
a 15-25 X 60-mm spotting scope. Coimts were made by
standing in a specific marked location and "mapping" the
cones on clear acetate. Changes in the "cone map" at each
observation were recorded, including the partial removal
of a cone. Cone count data were converted into seeds us-
ing an empirically derived value for the mean number of
seeds per cone (50.4 seeds/cone, Hutchins 1982). Partially
consimied cones were tallied by estimating from the ground
the percentage of seed remaining as described in Hutchins
and Lanner (1982). The seed harvest data were then plot-
ted against ciunulative time for both forest and meadow
sites. At each observation date samples of seeds were col-
lected to determine maturity and condition. The numbers
of filled, discolored, insect-attacked, and second-year-
aborted ovules were tallied. Mean dry weight of shelled
seeds, seed coat thickness, and caloric content of shelled
seeds were all obtained. Germination tests were also con-
ducted with seeds collected at 10 collection dates during
the 1980 field season.

Seedfall (caused by animal foraging) below the tree was
estimated by placing five, 1-m^ vnre mesh seed traps ran-
domly below trees. The tops of the seed traps were de-
signed to let seed fall through but to exclude rodents.
Data on the nimiber of cones in the tree, the area of the
tree crown, and the number of seeds falling per square
meter were used to estimate the magnitude of seed fall.

Predation on seed caches of whitebark pine was also
studied in 1979 and 1980. I simulated three types of
caches: (1) seed that lands on the soil surface from
foraging accidents; (2) seed cached at a depth of 3 cm

163

MONTANA

WYOMING

GRAND TEfoN
NAT.' PARk

I i ®

Surprise Lake ^

i A

SHOSHONE
NATIONAL
FOREST

LEGEND

— PAVED ROAD
® STUDY SITE

KILOMETERS

WYOMING

simulating a Clark's nutcracker (Nucifraga columbiana)
cache; and (3) seed cached 7 cm deep, which simulates
the most shallow red squirrel seed cache. Each individual
cache contained 10 seeds.

Animal time budget data were collected for the diurnal
species found foraging in the whitebark pine ecosystem.
Both quantitative and qualitative observations were made
on the various activities and behaviors of the animals; the
methods are detailed in Hutchins and Lanner (1982).

SEED DEVELOPMENT

Hutchins and Lanner (1982) monitored seed develop-
ment and found mean seed coat thickness and seed weight
to be significantly greater during the August 31 to
November 2 collecting period than prior to these dates.
Clark's nutcrackers were unable to extract whole seed
from a cone ;intil August 13 due to the thin, fragile seed
coats. The characteristic shell fragments of an unripened
"nutcracker cone" were left as evidence of their foraging
attempts. Thus, nutcrackers were unable to cache whole
seed until after this date.

By September 7-10, the cones had dried and turned a
dull brown from their previous moist, pitch-filled, purple
color. Whitebark pine cones are often referred to as inde-
hiscent; however, about 25 percent of the cones collected
after September 7 parted their scales slightly (4-8 mm;
n = 141). This still was not enough of an opening to allow
the seeds to fall out on their own accord.

Figure 1 — Locations of three study sites in Wyoming.

164

SEED CROP DEPLETION

Figure 3 (from Hutchins and Lanner 1982) shows seed
crop depletion followed a logistic curve (r^ = 0.99 in the
forest and = 0.96 in the meadow). Seed harvesting be-
gan somewhat earlier in the forest than in the meadow.
In fact, about 50 percent of the seed crop had been har-
vested in the forest by August 31, while in the meadow
this point was not attained until September 25. As seen
on the graph, no seeds remained from this mast year in
the forest stands by November 1. Only 0.1 percent of the
seeds still resided in the tree crowns of the meadow trees,
and these were no longer there when checked on 27 June,
1981.

Unless aided by an animal, cones of whitebark pine do
not fall off a tree to the ground except rarely during ex-
tremely large cone crops — as occurred in 1989 (Hutchins
and Lanner 1982; Lanner 1982; Mattson and Reinhart,
these proceedings). The following incident shows the ex-
tent Clark's nutcracker will go to acquire seed. Hutchins
and Lanner (1982) placed double-layer wire hardware
screening over several cones so there would be cones to
collect for germination studies. By late October — after
available whitebark pine seeds had become scarce — the
nutcrackers had ripped open the screening and harvested
the seed. Seed that does drop to the ground from foraging
accidents made up only 4.2 percent of the seed crop and
most of this seed (69 percent) was determined unviable
through examination of the contents in the shell during
1980 (Hutchins 1990).

ANIMAL INTERACTIONS WITH SEED

A variety of diurnal animals were observed active in or
imder whitebark pine trees. Many of these, however, were
never observed foraging on whitebark pine seed in this
study. These include: gray jay {Perisoreus canadensis).

common flicker (Colaptus auratus), Cassin's finch (Cora-
podacus cassinii), rosy finch (Leucosticte sp.), pine siskin
(Spinas pinus), dark-eyed junco (Junco hymelis), black
billed magpie (Pica pica), pine marten (Martes americana),
coyote (Canis latrans), and weasel (Mustela sp.).

Many feces of the mammal species were examined, but
no evidence of whitebark pine seeds was found. Gray jays
were commonly found in the crowns of whitebark pine but
were only observed hawking insects and caching fresh
carrion or boli in pine branches. Dow (1965) found this
species has little interest in pine seed during feeding tri-
als, although Turcek and Kelso (1968) describe its Eur-
asian cousin, the European jay (P. infaustus), as having
been observed taking and caching Siberian stone pine
(Pinus sibirica).

Of the birds, only the magpie could potentially be in-
volved in the dispersal of whitebark pine (Smith and Balda
1979). It is, however, very rare in the subalpine zone dur-
ing the maturation of whitebark pine seed (Hutchins and
Lanner 1982); consequently, any potential harvesting of
seed would be unimportant to the regeneration of white-
bark pine.

Many species have been observed foraging on whitebark
pine cones (Hutchins 1990, table 2). At this point I would
like to take a closer look at their role in the dispersal and
establishment of this pine.

BIRD DISPERSERS

Clark's Nutcracker — Nutcrackers were the most
common resident bird to visit the whitebark pine trees
(Hutchins and Lanner 1982). They were scattered about
whitebark pine stands in loose flocks foraging and caching
seed of this tree. They are dependent on these caches
year round (Giuntoli and Mewaldt 1978; Vander Wall
and Balda 1977; Vander Wall and Hutchins 1983).

August

September

October

November

Figure 3 — Seasonal course of
whitebark pine seed harvest by
vertebrates in Squaw Basin,
WY, in 1980.

165

Nutcrackers were observed harvesting seed as early
as July 13, 1979. During July, however, they appeared to
only be testing the cones for ripeness and primarily feed-
ing on the previous year's caches until mid-August. Seed
harvested at this time were lost to regeneration because
the seed coats were broken and the seed is ungerminable
vintil August 13. In 1980, the birds were able to success-
fully harvest whole, developed seed by August 15. Nut-
crackers harvested seed from cones at increasingly faster
rates (Hutchins and Lanner 1982) through early October,
when seed became hard to find (fig. 3). Nutcrackers were
never seen attempting to harvest Engelmann spruce or
lodgepole pine, even as the whitebark pine seed crop dwin-
dled in October. As the seed supply became scarce in mid-
October, I observed one bird checking over 50 cones for a
period of 613 seconds without finding a single seed.

By November 2, it appears nutcrackers were almost
totally dependent on their new seed caches, and would
remain so imtil the following August. These caches were
placed just below the soil surface (2-3 cm), and over 70 per-
cent of the caches were either one- two- or three-seed
caches (x = 3.2 + 2.8 seeds/cache; fig. 4). The largest cache
was 14 seeds. These data agree closely with Tomback
(1978) and Vander Wall and Balda (1977), although the
number of single-seeded caches was higher.

Clark's nutcrackers cached their seeds in almost every
conceivable soil type and microsite in the various study
sites. Specific examples were given in Hutchins (1990).
Although they may cache seeds on a wide variety of sites
in the Rocky Mountains, several studies have indicated
that south-facing slopes seem to be preferred (Lanner
1982; Lanner and Vander Wall 1980; Snethen 1980;
Tomback 1978; Vander Wall and Balda 1977). Seedling
establishment appears to be much more common on more
moist sites in the Rocky Mountains (Arno 1986; Arno and
Hoff 1989; Vander Wall and Hutchins 1983) than in the
Sierra Nevada (Tomback 1982).

40-

o
o

35
30
25
20
15
10

5 ^

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Number of Seeds per Cache

Figure 4 — Clark's nutcracker cache size fre-
quency. Data from 1 57 observations made at
several locations in Wyoming.

Transport distances varied greatly. Seeds were placed
as close as 50 m from the site of the harvested tree, or
were transported at least 3.5 km to the Breccia Cliffs on
the edge of the Squaw Basin site. With little seed left in
the trees by mid-October, the birds began retrieving many
of their caches made in the Squaw Basin meadows and
recaching them on the southwest-facing slopes of the
Breccia Cliffs.

Large flocks of nutcrackers would often cache seed
together. At Mount Washburn, a flock estimated at
150 birds was seen caching seed under an open-grown
whitebark pine stand, with 10-15 birds within a 10-m^
area. No aggression occurred among the birds during
these observations.

By November 2, 1980, snow covered many of the cach-
ing areas for the winter. Nutcrackers were observed on
several occasions successfully pecking through up to
25 cm of snow and ice to retrieve a cache during the win-
ter months. The Togwotee Pass area may receive up to
1,500 cm of snow a year, yet the windswept ridges and
south-facing slopes remain exposed enough for the birds
to retrieve their caches. Caches on the northeast-facing
slopes and under the forest canopy are more frequently
used during June, July, and August, as the snow recedes
from these sites last (Vander Wall and Hutchins 1983).

The number of seed an individual nutcracker caches
annually has been estimated in several studies. The
numbers for whitebark pine range from 32,000 in the
Sierras (Tomback 1982) to 98,000 in the Rocky Moun-
tains (Hutchins and Lanner 1982). Because of the num-
ber of variables that must be considered (such as flight
distances and the amount of available seed), these esti-
mates vary a great deal from site to site and year to year.

Studies by Tomback (1982), Vander Wall (1988), and
Vander Wall and Balda (1977), estimate that an indi-
vidual bird caches several times more seed than it needs
to survive through the winter and early spring. At that
time other food items become available — although the
nutcrackers continue to use their caches heavily until
the new cone crop begins to mature (Vander Wall and
Hutchins 1983). This leaves many unused caches for
potential germination and establishment of whitebark
pine seedlings (Hutchins and Lanner 1982).

Clark's nutcracker stands out from all the other poten-
tial seed dispersers in two major ways. First, it consis-
tently disperses seed in a way that increases the chance
of seedling establishment. The seed is placed just below
the soil surface and hidden from seed predators (Hutchins
and Lanner 1982; Lanner 1980).

Second, nutcrackers scatter caches across the
landscape — both long and short distances from the source
trees. Dispersal distances of up to 22 km have been ob-
served by Vander Wall and Balda (1977). Also, the scat-
tering of thousands of their caches reduces predation.

Steller's Jay (Cyanocitta stelleri) — Having many of
the behavioral and physiological attributes of the Clark's
nutcracker (Vander Wall and Balda 1981), the Steller's
jay could have potential as a dispersal agent of whitebark
pine. In this part of their range, they were found to be
primarily solitary foragers and uncommon visitors to the
whitebark pine forest. These birds did not forage for
whitebark pine seed until early September when the

166

cones dried and the scales separated slightly. Their bill
structure does not approach the efficiency of the nut-
cracker in prying apart cone scales to get to the seed
(Vander Wall and Balda 1981), thus they were often only
able to extract seed from cones that had been exposed by
nutcrackers. These jays harvested whitebark pine seed
from the ground and canopy in less than half of their ob-
served foraging time during the fall of 1980 and 1981
(Hutchins and Lanner 1982). The seed was always placed
between their feet and hammered with their bill to crack
the hull; nutcrackers primarily cracked the seed between
their mandibles.

Occasionally Steller's jays pouched their seed in their
elastic esophagus for later caching. They were observed
caching whitebark pine seeds on seven occasions; how-
ever, none of these were in the soil. The birds placed
the pine seeds in the crotch of a tree, a densely foliated
witches broom, and under dense lichen growth along a
tree branch. The largest number of seeds I observed be-
ing pouched at one time by this species was five, although
data from Vander Wall and Balda (1981) indicate these
birds could hold up to 32 seeds/pouchload. Because this
species does not cache whitebark pine seed in the soil and
has a seed foraging rate comparable to small passerines
(Hutchins and Lanner 1982), it is a very improbable agent
for seedling establishment.

Raven (Corvus corax) — third corvid rarely observed
foraging on whitebark pine seed was the common raven.
These birds had a great deal of difficulty extracting seed
from whitebark pine cones due to their thick, long bills.
They dropped most of the seed when foraging (Hutchins
and Lanner 1982). Caching of carrion was observed with
this species, but no observations were made of seed cach-
ing. In Asia, Reimers (1959 cited in Tiircek and Kelso
1968) described observations of ravens caching Japanese
{Pinus pumila) and Siberian stone pine (P. sibirica), al-
though it was not clear in what type of substrate.

Other Passerines — Pine grosbeaks (Pinicola enuclea-
tor) primarily foraged on whitebark pine seed that had
been exposed by Clark's nutcrackers breaking off cone
scales. Their large conical beaks also enabled them to
tear away cone scales to get at seed. Their foraging rates
were very slow compared to nutcrackers due to their tech-
nique and, as a consequence, they had little influence on
the depletion of the whitebark pine cone crop.

Grosbeaks are not known to cache food (Smith and
Balda 1979; Vander Wall 1990). During my observations
of this species, they cracked the seed coat and consumed
the seed in the tree crown. It is highly unlikely that they
covild pass an intact seed through their digestive tract.

The moxmtain chickadee (Parus gambeli) and red-
breasted nuthatch (Sitta canadensis) occasionally searched
through the whitebark pine cones during September after
the seeds were exposed by nutcrackers. Both these spe-
cies almost always dropped the seed from the tree crown,
as the seeds were too large for them to handle. Their un-
successful foraging bouts contributed to the small percent-
age of seed found on the soil surface, which was later con-
sumed by other animals. Neither bird species was observed
caching whitebark pine seed, although they are known to

cache smaller seeded conifer species imder bark (Smith
and Balda 1979).

MAMMAL DISPERSERS

Red Squirrel (Tamiasciurus hudsonicus) — Red

squirrels are common denizens of whitebark pine forests.
In fact they were the second most commonly observed ver-
tebrate next to the Clark's nutcracker (Hutchins and Lan-
ner 1982). These mammals actively defend their territo-
ries (Smith 1968), and like nutcrackers, harvest whitebark
pine cones and seeds as a fall and winter food source. Red
squirrels did not occiir on the meadow sites.

Red squirrels spent most of their foraging time (75.8 per-
cent) on whitebark pine cones or recovering dropped
whitebark pine seeds. This includes the time it takes
to place the cones in the midden. Most of the rest of the
observed foraging activities were collecting Engelmann
spruce cones (11.3 percent) and harvesting seeds of herba-
ceous plants (12.9 percent).

Red squirrels began harvesting cones as early as nut-
crackers harvested seeds (Jvily 13, 1979), but did so more
intensely during July while the nutcrackers relied more
heavily on the previous year's caches (fig. 3). Their forag-
ing rate is much higher than that of the other mammals
discussed in this paper because they usually harvest en-
tire cones. Most of the time, they remove the cone by pull-
ing it off the branch with their teeth, instead of cutting
the subtending branch (Hutchins 1990). Thus, cone re-
moval is likely to cause little change in tree growth form
or cone production.

Cone caching began on August 4 at Squaw Basin.
Red squirrels often would leave cones on the groimd below
the tree where they were cut for periods of up to 3 days
before caching them in their midden. Caching of other
conifer cone species began later with Engelmann spruce
on August 18 in 1980; subalpine fir (September 11 in
1980), and lodgepole pine (September 27 in 1980). All
cones were stored in "middens" {n = 114). These middens
can be quite extensive and are composed of many years of
cone debris above the soil surface (Finley 1969; Reinhart
and Mattson 1990; Smith 1970). Of the time spent cach-
ing food, 61.4 percent was devoted to whitebark pine
cones and 16.9 percent to whitebark pine seeds. Foraging
on the cones of other conifers in the subalpine made up
most of the rest of the caching time (7,304 seconds of ob-
servation). Red squirrels also cached some mushrooms
and herbaceous seeds.

Squirrels were also found to cache whitebark pine seeds
in their midden beginning aro\md September 16, 1980.
These seed caches were placed 6.5- to 40-cm deep {n = 6);
four of these observations were between 11 and 11.5 cm
deep. The number of seeds per cache ranged from 14 to
55 ix = 28.8; 19.2 seeds/cache; n = 4), although Kendall
(1981) foimd up to 176 seeds in a single hole. I examined
the seeds from two squirrel caches gind found all the seeds
sound.

Red squirrels actively chased Clark's nutcrackers from
the trees above their territories and their middens. On
two occasions, however, I watched nutcrackers steal whiter
bark pine seeds from squirrel middens.

167

Chipmunk (Eutamias sp.) — Chipmtinks are uncom-
mon visitors to whitebark pine tree crowns but do occur
on both open-meadow and forested sites. Most of their
time is spent on the ground, usually near cover plants like
sagebrush (Artemisia tridentata). They preferred herba-
ceous plant parts (lupine seed, grasses, etc.) on my subal-
pine study sites until the third week of September when
these plants died back and presimiably lost much of their
nutritive value. At that time, they clumsily foraged on
the remaining whitebark cone crop (-10 percent of the
seed was left in the forest, -40 percent in the meadow).
They more commonly foraged on the smsdl amount of seed
below the trees, which had been dropped by any of the
species discussed earlier. Tevis (1953) and Heller (1971)
similarly foimd that in the Sierra Nevada chipmxmks
devote little time to foraging on whitebark pine seed and
cones.

I found no evidence of chipmimks caching whitebark
pine seed. Broadbooks (1958) found the larder cache
depth of yellow pine chipmimks {Eutamias amoenus) to
average 28 cm. They use these types of caches as a winter
and spring food source when they are periodically aroused
during hibernation (Vander Wall 1990). I examined chip-
munk burrows to a depth of 20 cm without finding any
evidence of them harvesting and caching whitebark pine
seed.

Yellow pine chipmunks are known to scatter-hoard
smaller amounts of Jeffrey pine (Pinus jeffreyi) and pon-
derosa pine {P. ponderosa) seed in shallower caches from
1 to 25 mm in depth (Vander Wall 1992a). These species
of chipmunks in these studies, along with other rodents,
may retrieve many caches in the fall in the Sierra Nevada.
They use the seed to stock their winter larders, but
many seeds survive and germinate the following spring
(Vander Wall 1992b). Although chipmunks may play
a role in the afforestation of certain pines, there are no
data that support this conjecture for whitebark pine.
Vander Wall (1992a) provides strong evidence that Jeffrey
pine may arise from chipmunks (Tamias amoenus, specio-
sus, and quadramaculatus) scatter-hoarding seed. In
Japan, Hayashida (1989) found Siberian chipmunks
(T. sihiricus) were not an effective dispersal agent of
Korean stone pine {P. koraiensis) in a plantation. In this
study, the limited amount of seed these animals harvest
(see seed crop depletion section of this paper) apparently
precludes them from being of significance to whitebark
pine establishment in the Rocky Mountains. More detailed
observations of chipmunks in whitebark pine ecosystems
are needed to more accurately define their role as cache
predator and possible seed disperser.

Golden-Mantled Ground Squirrel (Spermophilus
lateralis) — Due to a slow foraging rate and relatively
rare occurrence in tree canopies, the golden-mantled
ground squirrel also consumes a limited amount of white-
bark pine seed. This species was found in a whitebark
pine tree only once in 188 hours of observation. It does
collect a limited amount of seed from the ground, which
falls from cones during foraging accidents involving other
animals.

Like chipmunks, groimd squirrels are known to scatter-
hoard shallow seed caches as well as make a deeper lar-
der hoard (MacClintock 1970; Vander Wall 1990). This

species also begins hibernation quite early in September
and would not have time to acquire much seed for storage.
As with the chipmimk, few studies have been performed
to define ground squirrel foraging and caching behavior.

SECONDARY DISPERSERS

Nocturnal Rodents — This group of animals, made up
primarily of mice and voles (family Cricetidae), must also
be considered as potential dispersal agents of whitebark
pine seed. They were not directly observed in this study,
but possible evidence of their foraging on whitebark pine
seed was discovered by the shelled seed left behind on my
simulated cache experiments.

Smface seed caches simulating the seed that was foimd
on the ground indicate it will not last long (Hutchins 1989;
McCaughey and Weaver 1990). Most of the seed in these
caches was consumed within 2 weeks after placement un-
der trees by evidence of the seed shells left behind. Al-
most all of the shells were left behind at the cache site,
indicating little if any caching was done.

There are two primary places where cricetid rodents may
obtain whitebark pine seeds: (a) from seed that falls to the
ground and (b) the discovery of animal (primarily nut-
cracker) caches. A small amount of seed (-4 percent of
the seed crop) was found to fall to the forest floor in the
mast year of 1980. About 69 percent of seed found on the
ground was determined imviable by examination of the
contents inside the shell. Although these rodents may
recache the seed, the seed was probably already placed
in a suitable site for seedling establishment by a Clark's
nutcracker. Thus, even if these cricetids relocate the seed
to another cache, their positive effect on the establish-
ment of whitebark pine is questionable at best.

Abbott and Quink (1970), working with the winged
eastern white pine {Pinus strobus) seeds, showed most
caches by these rodents were made less than 15 m from
the seed source. Thus, the habit of whitebark pine trees
pioneering open meadows and disturbed areas is highly
unlikely to arise from cricetid caches. Their study also
showed that of those caches not recovered by the time the
seed germinated, the germinated seedlings were soon con-
sumed by these animals. This information coupled with
the small amount of seed available on the ground for
these rodents suggests that they could rarely be respon-
sible for seedling establishment. Future studies need to
look at this group of dispersal agents more closely to fur-
ther delineate their role in whitebark pine ecosystems.

Mice of the genus Peromyscus are members of this
group of rodents, which is known to scatter-hoard pine
seedhngs (Abbott and Quink 1970; Vander Wall 1992a).
Like the chipmunks and ground squirrels, this species
searches for seeds on the ground and will either eat them
or cache them for later use. Because so few sound white-
bark seeds make it to the ground, it appears the role of
these small rodents in whitebark pine establishment is
insignificant.

Grizzly and Black Bears — Bears {Ursus arctos and
U. americana) primarily obtain seed from squirrel middens
(Kendall 1981, 1983; Mattson and Jonkel 1989), although
black bears are known to also break branches to harvest
the seed (Tisch 1961). Examinations of bear scat from

168

both species have yielded only a few intact whitebark pine
seeds (Hutchins and Lanner 1982). Even if these seeds
could germinate in the bear scats, they would produce an
insignificant number of whitebark pine seedlings that are
dispersed a very short distance fi'om the squirrel midden
the bear raided.

PERCENT OF SEED HARVESTED

On forested sites, about 63 percent of the seed was har-
vested by red squirrels and 36 percent by Clark's nut-
cracker. The other 1 percent was harvested by all other
animals combined (table 1).

In contrast, on the meadow sites, which have too few
trees for squirrels to exist, nutcrackers harvested almost
the entire crop of whitebark pine (99 percent). Details for
how these values were calculated are described in Hutchins
(1982, 1990).

In all but a year with a super-abundant cone crop, ani-
mals harvest nearly all of the seed crop by early November
in the Rocky Moimtains (Hutchins and Lanner 1982;
Vander Wall 1988). By this time, no cones remain on
the trees and seeds do not have a chance to survive long
enough on the grotmd £ind germinate as suggested by Day
(1967) and others. In this study I found that the seed and
cones that have been dropped to the groimd are eaten
within 3 weeks by various foragers.

CONCLUSIONS

Nutcrackers, squirrels, and the other whitebark pine
seed forager guilds will occasionally drop some seed to the
ground (~4 percent). Some seed is dropped intentionally
(Vander Wall and Balda 1977) and some by accident.
Most of this seed, however, is composed of unfilled or
aborted ovules (69.5 percent) and could never germinate.
Consequently, dispersal and establishment of whitebark
pine by small rodents like deer mice, ground squirrels,
and chipmunks would be too rare an event for a tree spe-
cies to survive and prosper.

We also cannot assert red squirrels to be an establisher
of whitebark pine seedlings. As I stated in earlier work,
I found no evidence of whitebark pine establishing on

squirrel middens (Hutchins 1989). The midden "seed bed"
is too deep in organic debris, and the activity on the midden
constantly distvu-bs the soil.

Squirrels do harvest most of the whitebark pine seed
in the forest, and they are the major seed predator on for-
ested sites. Their midden cone stores are also important
for providing grizzly and black bears with an important
fall food source (Reinhart and Mattson 1990).

Nutcrackers, on the other hand, are one of the most im-
portant biotic influences developing and changing subal-
pine communities. This species alone probably accoimts
for nearly all whitebark pine regeneration, except for
chance happenings. As far as the bird and the tree are
concerned, it is more profitable to cache in the open mead-
ow. Much less predation occurs on nutcracker caches in
the meadow (Hutchins 1990), and the small ridges are
usually fi'ee of snow due to wind action. Consequently,
the higher cache survival rate benefits tree regeneration
as well as the survival of the nutcracker. This more than
any factor may be why we see whitebark growing where
we do — pioneering the exposed ridges, roadside cuts,
burned sites, and meadow swales.

The open-grown meadow trees may be the most impor-
tant because the seeds from these trees will have a greater
chance of being dispersed by Clark's nutcracker. These
small tree islands lack the major seed predator (the red
squirrel). It should be added that Reinhart and Mattson
(1990) also foimd lower numbers of red squirrels using
pure whitebark pine stands. These also may be good
sources for increasing whitebark pine regeneration.

When nutcrackers forget where they placed a cache (for
example, Vander Wall 1982), or die, or a rodent does not
discover the seed cache — it has a chance to germinate.
By placing the seed in an excellent germination bed just
below the soil surface (2-3 cm) and also hiding the seed
fi'om easy discovery by seed predators, the bird may begin
a new forest stand.

Eventually, these whitebark pine trees modify the once-
open subalpine landscape so other more shade-tolerant
species such as Engelmann spruce and subalpine fir can
establish themselves in this commimity (Amo 1989;
Franklin and Dyrness 1973; Snethen 1980). The seed
and cover produced by whitebark pine attracts a large

Table 1 — Estimate of whitebark pine seed harvested by various animals in Wyoming

Mean seeds

Minutes spent

Foraging

Seeds

Number of

Percent seeds harvested

extracted/

foraging/

days/

harvested/

individuals

by all individuals

Species

minute^

day^

season

individual

visiting trees

Forest

Meadowr

Clark's nutcracker

7.9

180

129,402

448

36.3

99.4

Steller's jay

120

4,620

<.1

Raven

954

<.1

Noncorvids

120

4,704

Red squirrel

43.4

240

874,944

116

63.5

Chipmunk

1.7

120

7,140

<.1

'Seasonal average from observations made during August 15 to October 11, 1980.
^Estimate made from observed daily activity patterns.
'Meadow area lacks squirrels.

169

number of vertebrates (Amo and Hoff 1989; Hutchins
1990; Kendall and Arno 1990; Lonner and Pac 1990;
Tomback 1978).

SUMMARY

Whitebark pine is dependent on animal dispersal for
regeneration. Many animals interact with whitebark pine
seed, but only Clark's nutcracker was found to consis-
tently disperse seed in a way that leads to the regenera-
tion of this pine. These long-distance dispersal agents
should be the central focus in whitebark pine subalpine
community management.

ACKNOWLEDGMENTS

I thank Steve Vander Wall for his insight on small
mammals and pine seed dispersal. His comments were
a great help. Susan Hutchins, Ron Lanner, and Diana
Tomback helped with earlier versions of this manuscript.
I thank them for their efforts.

REFERENCES

Abbott, H. G.; Quink, T. F. 1970. Ecology of eastern white
pine seed caches made by small mammals. Ecology.
51: 271-278.

Arno, S. 1986. Whitebark pine cone crops — a diminishing
source of wildlife food? Western Journal of Applied For-
estry. 1(3): 92-94.

Arno, S.; Hoff, R. J. 1989. Pinus albicaulis Engelm.:
whitebark pine. In: Burns, R. M; Honkala, B. H., tech.
coords. Silvics of North America. Vol. 1. Conifers. Agric.
Handb. 654. Washington, DC: U.S. Department of Agri-
culture, Forest Service: 269-279

Broadbooks, H. E. 1958. Life history and ecology of the
chipmunk {Eutamias amoenus) in eastern Washington.
Misc. Publ. 103. Ann Arbor, MI: Museum of Zoology,
University of Michigan: 5-42.

Day, R. J. 1967. Whitebark pine in the Rocky Mountains
of Alberta. Forestry Chronicle. 43: 278-282.

Dow, D. D. 1965. The role of saliva in food storage by the
gray jay. Auk. 82: 139-154.

Finley, R. B., Jr. 1969. Cone caches and middens ofTam-
iasciurus in the Rocky Mountain region. Misc. Publ. 51.
In: Contributions in mammalogy. Lawrence, KS: Uni-
versity of Kansas, Museum of Natural History: 233-274.

Franklin, J. F.; Dyrness, C. T. 1973. Natural vegetation
of Oregon and Washington. Gen. Tech. Rep. PNW-8.
Portland, OR: U.S. Department of Agriculture, Forest
Service, Pacific Northwest Forest and Range Experi-
ment Station. 417 p.

Giuntoli, M.; Mewaldt, L. R. 1978. Stomach contents
of Clark's nutcracker in western Montana. Auk.
95: 595-598.

Hayashida, M. 1989. Seed dispersal by red squirrels and
subsequent establishment of Korean pine. Forest Ecol-
ogy and Management. 28: 115-129.

Heller, H. C. 1971. Altitudinal zonation of chipmunks
{Eutamias): interspecific aggression. Ecology.
52(2): 312-319.

Hutchins, H. E. 1982. The role of birds and mammals
in the dispersal and establishment of whitebark pine.
Logan, UT: Utah State University. 67 p. Thesis.

Hutchins, H. E. 1990. Whitebark pine seed dispersal and
establishment: who's responsible? In: Schmidt, W. C;
McDonald, K. J., comps. Proceedings — symposiimi on
whitebark pine ecosystems: ecology and management of
a high-mountain resource; 1989 March 29-31; Bozeman,
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart-
ment of Agriculture, Forest Service, Intermountain Re-
search Station: 245-255.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Kendall, K. C. 1981. Bear use of pine nuts. Bozeman, MT:
Montana State University. 27 p. Thesis.

Kendall, K. C. 1983. Use of pine nuts by grizzly and black
bears in the Yellowstone area. Interagency conference
on bear research and management. In: Bears — their
biology and management. Calgary, AB: International
Association for Bear Management. 5: 166-173.

Kendall, K. C; Arno, S. F. 1990. Whitebark pine— an im-
portant but endangered wildlife resource. In: Schmidt,
W. C; McDonald, K. J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and manage-
ment of a high-mountain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 264-273.

Lanner, R. M. 1980. Avian seed dispersal in the ecology
and evolution of limber and whitebark pines. In: Sixth
North American forest biology workshop proceedings.
Edmonton, AB: University of Alberta: 14-48.

Lanner, R. M. 1982. Adaptations of whitebark pine for
seed dispersal by Clark's nutcracker. Canadian Journal
of Forest Research. 12(2): 391-402.

Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of lim-
ber pine seed by Clark's nutcracker. Joiirnal of For-
estry. 78(10): 637-639.

Lonner, T. N.; Pac, D. F. 1990. Elk and deer use of white-
bark pine forests in southwestern Montana: an ecologi-
cal perspective. In: Schmidt, W. C; McDonald, K. J.,
comps. Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-mountain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
237-244.

MacClintock, D. 1970. Squirrels of North America. New
York: Van Nostrand Reinhold. 184 p.

McCaughey, W. W.; Weaver, T. 1990. Biotic and micro-
site factors affecting whitebark pine establishment. In:
Schmidt, W. C; McDonald, K. J., comps. Proceedings —
symposivun on whitebark pine ecosystems: ecology and
management of a high-mountain resource; 1989
March 29-31; Bozeman, MT. Gen. Tech. Rep. INT-270.
Ogden, UT: U.S. Department of Agriculture, Forest Ser-
vice, Intermountain Research Station: 140-150.

Reinhart, D. P.; Mattson, D. 1990. Red squirrels in the
whitebark zone. In: Schmidt, W. C; McDonald, K. J.,

170

comps. Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-mountain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
256-263.

Smith, C. C. 1968. The adaptive nature of social organiza-
tion in the genus of tree squirrels Tamiasciurus. Eco-
logical Monographs. 38: 31-63.

Smith, C. C. 1970. The coevolution of pine squirrels
{Tamiasciurus) and conifers. Ecological Monographs.
40: 349-371.

Smith, C. C; Balda, R. P. 1979. Competition among in-
sects, birds, and mammals for conifer seed. American

Zoologist. 19: 1065-1083.
Snethen, K. L. 1980. Whitebark pine {Pinus alhicaulis)

invasion of a subalpine meadow. Logan, UT: Utah State

University. 76 p. Thesis.
Tevis, L., Jr. 1953. Stomach contents of chipmunks and

mantled ground squirrels in northeastern California.

Journal of Mammalogy. 34: 316-324.
Tisch, E. L. 1961. Seasonal food habits of the black bear

in the Whitefish Range of Northwestern Montana.

Missoula, MT: University of Montana. 67 p. Thesis.
Tomback, Diana F. 1978. Foraging strategies of Clark's

nutcracker. The Living Bird. 16: 123-161.
Tomback, Diana F. 1982. Dispersal of whitebark pine

seeds by Clark's nutcracker: a mutualism hypothesis.

Journal of Animal Ecology. 51: 451-467.

Turcek, F. J.; Kelso, L. 1968. Ecological aspects of food
transportation and storage in the Corvidae. Commimi-
cations in Behavioral Biology. 1: 277-297.

Vander Wall, S. B. 1982. An experimental analysis of
cache recovery in Clark's nutcracker. Animal Behaviour.
30: 84-94.

Vander Wall, S. B. 1988. Foraging by Clark's nutcracker
on a rapidly changing pine seed source. Condor. 90:
621-631.

Vander Wall, S. B. 1990. Food hoarding in animals.
Chicago, IL: University of Chicago Press. 445 p.

Vander Wall, S. B. 1992a. Establishment of Jeffrey pine
seedings from animal caches. Western Jovirnal of Ap-
plied Forestry. 7(1): 14-20.

Vander Wall, S. B. 1992b. [Personal communication].
July. Reno, NV: University of Nevada, Department
of Biology.

Vander Wall, Stephen B.; Balda, R. P. 1977. Coadapta-
tions of Clark's nutcracker and the pinon pine for effi-
cient seed harvest and dispersal. Ecological Monographs.
47: 89-111.

Vander Wall, Stephen B.; Balda R. P. 1981. Ecology
and evolution of food-storage behavior in conifer-seed
caching corvids. Zeitschrift fiir Tierpsychologie. 56:
217-242.

Vander Wall, Stephen B.; Hutchins, Harry E. 1983.
Dependence of Clark's nutcracker, Nucifraga Colum-
biana, on conifer seeds during the postfledging period.
The Canadian Field Naturahst. 97(2): 208-214.

171

SIZE OF PINE AREAS IN RELATION
TO SEED DISPERSAL

Hermann Mattes

Abstract — Pine seeds are either wind- or bird- (animal-)
dispersed. Only well-adapted seeds support large ranges, which
are of similar size in both types of dispersal. Seeds adapted to
wind dispersal should have a ratio of seed weight to wing length
below 1.8 mg/mm. Bird-dispersed seeds have a weight between
90 and 550 mg. Species with seeds above 100 mg but without
zoochorous features have small or even relictic ranges. Large
seeds are advantageous under severe conditions and in competi-
tion. Consequently, the bird-pine mutualism contributes to
larger ranges of those pine species.

Distribution ranges of plant and animal species are de-
termined by four main factors: ecological potency of the
species; effectiveness of seed dispersal; size of area avail-
able; and time available. Ecological studies on plants and
vegetation mostly focus on site conditions. Indeed, compe-
tition for nutrients, water, light, and so on most often
turn out to be the main factor limiting distribution. How-
ever, first of all, seeds must be able to reach the site in
question. Therefore, seed dispersal is an important factor
in plant ecology. Patchy and irregular patterns of species
on ruderal or fallow ground are a result of dissemination.
Otherwise vegetation history in the Holocene gives evi-
dence that time since last glaciation was not long enough
for full expansion of all species. We expect effectiveness
of seed dispersal to be correlated with rapidness of expan-
sion and with size of range.

Small, winged seeds are easily disseminated by wind,
which is present almost everywhere and emytime. Com-
petition of the seedling within a closed vegetation cover
as well as harsh environmental conditions require large
seeds with a high amount of nutrient reserves. Many
pine species have developed an almost obligate mutualis-
tic relationship to birds (Nucifraga, Gymnorhinus) for
seed dispersal. Pines with bird-dispersed, large seeds be-
came dominating tree species at arctic and alpine timber
lines, in xeric environments as well as in mesic lowland
forests.

The genus Pinus with more than 100 species provides
a good chance to test effectiveness of seed dispersal. All
pine seeds are of a rather similar structure; the most obvi-
ous differences are seed size and wing length. The pri-
mary type is very likely an anemochorous seed with a
long wing. Second, many species have developed large

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Hermann Mattes is Professor, Institut fur Geographie, Abteilung
Landschaftsokologie, Westfalische Wilhelms-Universitat, Robert-Koch-
Str. 26, W-4400 MOnster, Germany.

seeds with shortened or missing vnngs. These seeds are
disseminated mainly by birds, or occasionally by squirrels.
Apart from that, seeds and cones show many morphologi-
cal and phenological adaptations for dispersal, which are
discussed elsewhere (Lanner 1980, 1982; Vander Wsdl and
Balda 1977).

In this paper some relations of seed characteristics and
size of ranges are discussed. The main hjrpothesis is to
find the larger ranges within pine species well adapted
to seed dispersal.

MATERIALS AND METHODS

From 65 out of about 106 pine species in total enough
detailed information is available for a comparison of dis-
persal abilities and range sizes. The pine species are
numbered following Mirov (1967).

Distribution ranges are of a complex nature. Effective-
ness of dispersal would be measured best by the propor-
tion of the potential area that really has been occupied by
the species. Unfortunately, we do not know exactly the
potential ranges of almost all plant species. An interest-
ing idea would be to measure the distances or Eireas occu-
pied after the end of the last glaciation. However, refugials
and paths of dispersal are very incompletely known.
What we can clearly see are area and distances within
ranges. However, it seems not to be adequate to use the
area of a range. It is highly influenced by size of the re-
gion (the continent or vegetation zone that is inhabited
by the species concerned). Error should be less using the
distance between the outermost points of a range. This
was calculated as the orthodrome distance.

Split or disjunct ranges have been treated in the same
way as continuous ranges. It has proved to be too difficult
to decide whether an interrupted range was due to natu-
ral factors or to hxmian influences, especially in regions
that were densely populated for a long time such as East
Asia and the Mediterranean.

In the present study all 105 pine species listed in Mirov
(1967), and in addition Pinas longaeva, are considered.
Subspecies have not been considered. Reinge maps £ind
information was taken from Kriissmann (1968), Little
(1971), Meusel and others (1965), and Mirov (1967).

In only a few experiments flying ability of pine seeds
was examined (Lanner 1985; Muller-Schneider 1977).
To estimate dispersal abilities of numerous species, indi-
rect methods have been used. Simple, but reliable meas-
lu-ements for anemochorous seeds are weight of cleaned
seeds and wing length. Such data are available for most
of the species (Kriissmann 1972; USDA 1974). The ratio
of seed weight to wing length is used as an indicator for
potency of dispersal. Volume and surface of the seed and
shape and surface of the vring would be of great interest,

172

but these data are not available for most species. Zooch-
orous species are even more difficult to evaluate. Seed
and cone characteristics of species fully developed for bird
dispersal are compared morphologically with less-adapted
species.

DISPERSAL OF PINE SEEDS

Seed dispersal either by wind or by birds has caused
specific adaptions. Most of the pine species can be attrib-
uted to these two main types. A narrow seed-weight to
wing-length ratio is considered to be favorable to flight
ability. From figure 1 with 46 pine species Usted we can
assume:

• In general, the length of seed wings increases until
seed weight reaches about 90 mg (P. palustris: 93 mg/
36 mm); however, the ratio of seed weight to wing length
increases obviously with increasing seed weight.

• Winged seeds of more than about 100 mg have an
unfavorably high ratio of seed weight to wing length; wing
length is insvifficient even in species with huge cones and
long cone scales. These seeds are expected to have no
adequate flight abilities.

• The smallest seed weight of a wingless seed species
{P. flexilis) is 93 mg; most of the species with heavier
seeds are wingless, or seed wings separate easily (P. penta-
phylla No. 83, P. himekomatsu No. 84, P. strobiformis
No. 35; also in P. pinea No. 73).

Comparing morphological features with area size we
can conclude the following (fig. 2, table 1).

• Wind-dispersed pine species with a low ratio of seed
weight to wing length often have larger distribution
ranges than species with a high ratio.

• There is evidence for a log-linear relationship for the
largest ranges of each class of seed weight/wing length ra-
tio. Species with a ratio higher than 2 mg/mm are less
distributed on an average. It is supposed that dispersal
abilities limit the range of distribution. A remarkable ex-
ception is P. pinea (No. 73). Its origin is not known, and
it has been cultivated in many places since Phoenicean
times, approximately 4,000 years ago. Therefore, the
whole present range had to be taken into consideration.

• Many species do not reach an area as large as could
be expected in view of their theoretical abilities for seed
dispersal. This can be partly explained by geographical
and ecological reasons. Species of southeastern North
America (Nos. 22-31) depend on a relatively small area
limited by the ocean or woodless plains. The range of
Cahfomian pines (Nos. 5, 6, 17-19) of the Mediterranean
type is limited by climate. Mexican pines (seed data only
for P. chihuahua and patula, Nos. 40 and 61) also are re-
stricted geographically and ecologically to narrow moxm-
tainous belts. Pinus peuce and P. heldreichii (Nos. 71 and
75) are endemic to the Balkan mountains in southeastern
Europe.

• Ranges of bird-dispersed pines are of similar size as
those of wind-dispersed pines; differences are mainly due
to the very restricted range of the bird-dispersed P. quad-
rifolia in America and the huge range of wind-dispersed
P. sylvestris in Eurasia. Some bird-dispersed pine ranges
are among the largest of all pines. Concerning size of

ranges, dispersal by birds (and sqmrrels) is as efficient
as dispersal by wind.

• Pine ranges in America are smaller than those in
Eurasia. This is an effect of the larger land mass in the
Old World hemisphere. Only species with well-adapted
dispersal mechanisms are affected; species with relicted
ranges show no difference in area size.

GENUS PINUS DISTRIBUTION

Most of the pine species are holarctic. Pines occupy
a wide variety of sites of which the most extreme are
located at timber lines in cold and arid climates as well
as in very dry or perhvunid regions in the subtropics and
tropics. Many of the pine species are restricted to specific
sites. Thus, an effective dispersal agent is needed to
reach suitable sites.

Recent centers of species diversity (fig. 3) are Mexico
and the southwestern States of the U.S.A. Secondary cen-
ters of diversity are eastern North America, eastern Asia,
and the Mediterranean region. These were refugial areas
during glaciation. Although none of these areas is neces-
sarily the origin place of the genus Pinus, recent regions
of high species numbers of pines are assumed to be places
where pines have existed for a long time.

In contrast, there are large areas populated only by a
few pine species. This is the case especially in the present
boreal zone. Also, we find relatively few pine species in
Central America south of Tehuantepec, on the Caribbean
Islands, in the Himalayas, and in southeastern Asia.
Probably, these areas have been occupied by pines for
only a relatively short time.

The boreal zone was reoccupied by pines in the postgla-
cial time, and we know that some species are stiU expand-
ing (for example see Gorchakovsky 1993, for Pinus sibi-
rica). Nevertheless, their ranges belong to the largest
ranges of pine. All boreal pine species are among the spe-
cies best adapted either for wind dispersal (figs. 1 and 2;
Nos. 16, 20, 21, 32, and 69) or bird dispersal (Nos. 67 and
68). Wind-dispersed pines have excellent seed-weight/
wing-length ratios from 0.4 to 1.1 mg/mm.

Data are available for only a few species of southern
marginal regions (P. wallichiana, No. 91; P. merkusii,
No. 101; P. khasya, No. 104). With a ratio of seed weight/
wing length of 1.2 to 1.7 mg/mm, adaptation for wind dis-
persal in these species is relatively good.

Species that were more widely distributed in the Ter-
tiary or early Pleistocene, and that now occupy relictic
ranges only, show very different ratios of seed weight/
wing length. For P. aristata (No. 5; 1.2 mg/mm) and per-
haps P. balfouriana (No. 6; 2.1 mg/mm) and P. longaeva
(No. 6a) seed dispersal might not be a restricting factor.
This svirely is the case in P. torreyana (No. 11).

Species with reduced seed wings have less chances to
be dispersed by wind. Some of them are on the way to
developing zoochorous features. An advanced state of
zoochory shows in especially P. strobiformis (No. 35), and
somewhat less in P. pentaphylla (No. 83) and P. himeko-
matsu (No. 84) (Hayashida 1989); they have large seeds
and wings that easily separate fi-om the seed. These spe-
cies are already mainly disseminated by birds and ani-
mals. Pinus peuce (No. 71) has a small seed (40 mg) and

173

length of seed wing

0.4
0.4

0.5 n

0.4 ]
0.4 D
0.7
0.6
0.S
0.6 h
0.7]
1.6 3

1.13
1.1 J

0.7 1
1.5 □

16
5.9

n 84

seed weight

banks iana 32
contorta 16
sylvestris 69
clausa 28
serotina 30
pa tula 61
resinosa 21
virginiana 27
chihuahua 40
pungens 31
halepensis 77
strobus 20
nigra 74

1.0 p monticola 4
attenuata 17
heldreichii 75

1.3 p merkusii 101

2.1 □ aristata 6

1.3 ZJ elliottii 23
1.5 tzi radiata 19

ponderosa 14
peuce 71
engelmannii
pinaster 80
wall ichi ana

□ palustris

□ flexilis 2

□ pumila 6 8
m! pentaphylla 83
~i jeffreyi 13

:l strobiformis 35
I albicaulis 1
1 lamber tiana 3

1.3
5.7

1.6 i

1.9 ZD

1.7
2.6

7.7

[j378
56

91
22

181

cembra 7 0
edulis 8
j sibixica 67

U armandii 86
i coulteri 12

. 1 monophylla 7

1 gerardiana 92
" I cembroides 36
• ^ I quadrifolia 9

: "~1 koraiensis 82

I pinea 7 3
: . ' I sabiniana

torievana] ll

[mml 30 20

0 0 100 200 300 400 500 600 700 800 900 [mgl

Figure 1— Wing length (mm, left), weight (mg, right), and ratio of wing length to weight (mm/mg,
middle) of seeds of 46 pine species. Numbers following species names according to Mirov (1967).
For sources see text. Species are arranged according to seed weight.

is not very attractive for nutcrackers, although it is used
for hoarded food sometimes as is P. longaeva (Lanner
1988). It is assumed that dispersal by nutcrackers is of
some relevance in establishing these two species at tim-
berline. Ranges of both species are small.

Other pine species such as P. lambertiana, P. sabiniana,
P. torreyana, P. coulteri, and P. jeffreyi (Nos. 3, and 10-13)
have large seeds but lack further adaptations for disper-
sal by birds and animals. Since these species have no ef-
fective dispersing agent, their ranges are relatively small.

174

{mg/mmj in
403 6-j

^ 148
"5

55
5 7.4

2.7
1.0

5 -1
4
3 -
2 H

1
0

0.4 -1 -i
0.6 -1.5

•63

7 36 1 86
ftX'n ** **** • ■ *

92 8 70 35 2 8 2 6 7 6 8

' 73

• 83

'12

• 3

• 13

• Ca

. 53

-91

• 78
•22

5 '19, 24';i-0l'77
23 1 a . • • •14

40.,7.95 'ZS^^

.28 26 ^52^ .32
•l6

•69

5.5
246

6.0
404

6.5
685

7.0 7.5 8.0 8.5 9.0 9.5

1097 1808 2961 4815 8103 13360 [km]
size of range

Figure 2 — Size of ranges in relation to the ratio wing length/
seed weight of 44 wind-dispersed pines; for comparison,
range size of 13 bird-dispersed pine species on top of the
figure. For code numbers for pine species see appendix;
for sources see text.

PiniLs pinea, as mentioned earlier, is an exception because
it has been cultivated throughout history.

CONCLUSIONS

Competition in seedlings causes a high selective pres-
sxire for large seeds. Even among wind-dispersed pines
seed size tends to enlarge despite a strong negative effect
on dispersal ability. Obviously, at 90 to 100 mg seed
weight dispersal success by wind has diminished so far
that other agents are essential. From 24 haploxylon pine
species, 19 species have wingless or almost wingless and

large seeds, which are proven to be or expected to be bird
dispersed. There is no doubt that zoochory has developed
in parallel in several sections of haploxylon pines (Lanner
1989; Tomback and Linhart 1990). Three diploxylon spe-
cies (P. torreyana, No. 11; P. sabiniana, No. 12; P. pinea,
No. 73) have large, nearly wingless seeds but lack further
adaptations for zoochory.

Range sizes of bird-dispersed pines are not obviously
different from those of wind-dispersed pines. Dissemina-
tion of pine seeds by birds is at least as effective as wind
dispersal. The interaction of pines and nutcrackers or
jays is well balanced. Even scattered stands of zoochorous

Table 1 — Average size of ranges (km distance) of pine species with different dispersal agents. Wind-dispersed
pines are arranged according to their seed characteristics (seed weight/wing length in mg/mm)

Distribution ranges

Pine species dispersed by: America Eurasia Global

Birds

1,557

= 7)

2,680

in =

2,075

= 13)

Wind (0.4-0.8 mg/mm)

1,913

= 13)

M,247

in =

2,351

= 16)

Wind (0.9-2.0 mg/mm)

1,926

= 9)

2,610

in =

2,225

= 16)

Wind (>2.0 mg/mm)

1,121

= 7)

21,055

in =

21,097

= 11)

Species considered above

1,693

= 36)

22,566

in =

20)

21,994

= 57)

All species

1,413

= 67)

1,779

in=

39)

1,548

= 106)

'Without P. sylvestris, average range size is 2,150 km.
^Numbers without P. pinea.

175

176

pine trees support all needs for territories of nutcrackers
concerning breeding, feeding, overwintering, and social
life. Nutcrackers and jays have still preserved character-
istic euryoecious features of corvids, especially omnivory.
In Eurasia, where there is almost no other animal com-
peting with the nutcracker for seed hoarding, the range
of the nutcracker surpasses the ranges of zoochorous
pines in many regions.

Within wind-dispersed pines we find that large-sized
ranges suppose small seeds with a ratio of weight to wing
length below approximately 1.8 mg/mm, but not vice-
versa because of geographical barriers or specialized fea-
tures of a species. However, pine species with a very high
ratios of seed weight/wing length (above 1.8 mg/mm) al-
ways have restricted areas.

Altogether, pine seed dispersal by birds was very suc-
cessful dvuing evolution. Pine species became able to oc-
cupy additional and more different sites, as well as under
extreme conditions at the borders of tree growth as under
mesic conditions. Consequently, bird-dispersed pines
were able to enlarge their ranges.

REFERENCES

Critchfield, William B.; Little, Elbert L., Jr. 1966. Geo-
graphic distribution of the pines of the world. Misc.
Pub. 991. Washington, DC: U.S. Department of Agricul-
ture, Forest Service. 97 p.

Gorchakovsky, P. L. 1993. Distribution and ecology of Si-
berian stone pine (Pinus sibirica) in the Urals. (These
proceedings).

Hayashida, Mitsuhiro. 1989. Seed dispersal and regenera-
tion patterns of Pinus parviflora var. pentaphylla on
Mt. Apoi in Hokkaido. Research Bulletin of the College
Experiment Forests, Faculty of Agriculture, Hokkaido
University. 46(1): 177-190.

Kriissmann, Gerd. 1968. Die Baume Europas. Berlin and
Hamburg: Parey Verlag. 140 p.

Kriissmann, Gerd. 1972. Handbuch der Nadelgeholze.
Hambiu-g und Berlin: Parey- Verlag.

Lanner, Ronald M. 1980. Avian seed dispersal as a factor
in the ecology and evolution of limber and whitebark

pines. Proc. 6th North American Forest Biology Work-
shop, University of Alberta, Edmonton, Canada: 15-48.

Lanner, Ronald M. 1982. Adaptations of whitebark pine
for seed dispersal by Clark's nutcracker. Canadian
Journal of Forest Research. 12(2): 391-402.

Lanner, Ronald M. 1985. Effectiveness of the seed wing of
Pinus flexilis in wind dispersal. Great Basin Naturalist.
45(2): 318-320.

Lanner, Ronald. 1988. Dependence of Great Basin bristle-
cone pine on Clark's nutcracker for regeneration at high
elevation. Arctic and Alpine Research. 20(3): 358-362.

Lanner, Ronald M. 1989. Biology, taxonomy, evolution,
and geography of stone pines of the world. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposium on vv^hitebark pine ecosystems: ecology and
management of a high-mountain resource. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
14-24.

Little, Elbert L., Jr. 1971. Atlas of United States trees.

Vol. 1: Conifers and important hardwoods. Misc. Publ.

1146. Washington, DC: U.S. Department of Agriculture.

9 p. and 200 maps.
Meusel, H.; Weinert, E.; Jager, E. 1965. Vergleichende

Chorologie der zentraleuropaischen Flora. Bd. I. Jena:

Fischer- Verlag. 9 p. and 200 maps.
Mirov, Nicholas T. 1967. The genus Pinus. New York:

Ronald Press. 602 p.
Miiller-Schneider, Paul. 1977. Verbreitungsbiologie

(Diasporologie) der Blutenpflanzen. Veroffentlichungen

des geobotanischen Institutes der ETH Zurich, Stiftung

Riibel. Vol. 61. 226 p.
Tomback, Diana F.; Linhart, Yan B. 1990. The evolution

of bird-dispersed pines. Evolutionary Ecology. 4: 185-219.
U.S. Department of Agriculture, Forest Service. 1974.

Seeds of woody plants in the United States. Agric.

Handb. 450. Washington, DC. 838 p.
Vander Wall, Stephen B.; Balda, Russ P. 1977. Coadapta-

tions of the Clark's nutcracker and the pinyon pine for

efficient harvest and dispersal. Ecological Monographs.

Durham. 47(1): 27-37.

177

APPENDIX: CODING LIST OF NUMBERS OF PINE SPECIES CONSIDERED
(AFTER MIROV 1967) wi^c»ii^r.rtr.u

Western North America

Haploxylon pines

1 albicaulis

2 flexilis

3 lambertiana

4 monticola

5 balfouriana

6 aristata
6a longaeva

7 monopylla

8 edulis

9 quadrifolia

Diploxylon pines

10 sabiniana

11 torreyana

12 coulteri

13 jeffreyi

14 ponderosa

16 contorta

17 attenuata

18 muricata

19 radiata

Eastern North America

Haploxylon pines

20 strobus

Diploxylon pines

21 resinosa

22 palustris

23 elliotti

24 toeda

25 echinata

26 glabra

27 virginiana

28 clausa

29 rigida

30 serotina

31 pungens

32 banksiana

Mexico

Haploxylon pines

35 strobiformis

36 cembroides

Diploxylon pines
40 chihuahua
53 engelmanni

Northern Eurasia

Haploxylon pines

67 sibirica

68 pumila

Diploxylon pines

69 sylvestris

Mediterranean Region

Haploxylon pines

70 cembra

71 peuce

Diploxylon pines

73 pinea

74 wi^m

75 heldreichii

76 montana i=mugo)

77 halepensis

78 brutia
80 pinaster

Eastern Asia

Haploxylon pines

82 koraiensis

83 pentaphylla

84 himekumatsu
86 armandi

91 wallichiana (=griffithi)

92 gerardiana

Diploxylon pines
95 densiflora
101 merkusii
104 khasya

178

THE REGENERATION PROCESS
OF WHITEBARK PINE

Ward W. McCaughey

Abstract — ^Whitebark pine (Pinus albicaulis Engelm.) regener-
ates similarly to European and Asian stone pines. Our knowl-
edge of this process is essential because whitebark pine is impor-
tant to the North American grizzly bear {Ursus arctos horribilis),
other wildlife species, hydrology of high-elevation ecosystems,
and esthetics. This paper simimarizes available information on
the whitebark pine regeneration process, beginning from bud ini-
tiation through germination and seedling survival. Major factors
limiting germination and regeneration success are discussed. We
continue to gain knowledge about the regeneration process of
whitebark pine, but further research is needed to fully imder-
stand delayed germination mechanisms and habitat require-
ments for optimimi regeneration success.

Whitebark pine (Pinus albicaulis Engelm.) is an impor-
tant food source for the endangered grizzly bear (Ursus
arctos horribilis), red squirrels (Tamiasciurus hudsonicus),
the Clark's nutcracker (Nucifraga columbiana Wilson),
and a multitude of other birds and mammals (Hutchins
and Lanner 1982; Kendall 1983; Knight and others 1987).
It is also significant as a hydrologic stabilization plant
and an esthetic feature of high-elevation communities
(Schmidt and McDonald 1990). It provides important
cover for wildlife and is used as an ornamental for land-
scaping but has only minor significance as a timber pro-
ducing species (McCaughey and Schmidt 1990).

General information is available and is summarized
in this paper on the natiu-al regeneration process of white-
bark pine, but more specific facts are needed for efficient
management of this North American stone pine (Amo and
Hoff 1989; McCaughey and Schmidt 1990). This paper
describes the regeneration process beginning firom bud
initiation throiigh germination and seedling survival.

CONE AND SEED DEVELOPMENT

Whitebark pine cone and seed development begins vnth
cone initiation and ends with seed matiiration. Climatic
conditions influence the timing of development and matu-
ration of cone initiation and seed maturation.

Cone Initiation

The process of cone initiation and development of stami-
nate and ovulate strobili for whitebark pine has not been

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Ward W. McCaughey is a Research Forester, Intermountain Research
Station, Forest Service, U.S. Department of Agriculture, Bozeman, MT
59717-0278.

specifically studied. Mirov (1967) and Krugman and
Jenkinson (1974) provided generalized descriptions of the
initiation process for the genus Pinus. Female and male
cone initiation of whitebark pine probably occiu- during
or just prior to winter bud formation from mid-August to
mid-September (Schmidt and Lotan 1980).

Cone Development

Female and male buds overwinter and begin further
development and growth in April and May depending on
elevation. In 1992, staminate cones matured and shed
pollen during Jxme and early Jvdy at high elevations
(2,500-2,600 m) in southwestern Montana (McCaughey
1992) and in May and June at lower elevations (1,550-
1,650 m) in northern Idaho (Hoff 1992). Mature male
cones are about 1 cm wide by 1 cm long and female cones
are 1 cm wide by 2 cm long during early development
when they are pollen receptive. After pollination female
cones grow to a first-year size of about 3 cm long by 2 cm
wide. Female cones remain on the tree, while male cones
fall off after pollen dispersal. Pollination occurs from late
Jime to late Jvdy, and fertilization occxirs about 12 to 13
months later, simultaneous with rapid cone enlargement
during June and July (Krugman and Jenkinson 1974).

Seed Maturation

Following fertilization the embryo develops and dif-
ferentiates into cotyledons, plumule, and radicle (Mirov
1967). Whitebark seeds reach maturity between mid-
August and mid-September depending on elevation and
climatic conditions during the ripening process (Krugman
and Jenkinson 1974; McCaughey, in press a). The entire
process from cone initiation to cone and seed matiuity
takes about 24 months.

Cone and Seed Insects and Diseases

Whitebark pine cones and seeds are exposed to insects
and disease that reduce cone and seed siirvival during the
2 years of development (Edwards 1990). Cone and seed
insects that damage whitebark pine are cone worms (Dio-
ryctria spp. and Eucosma spp.), cone beetles (Conophtho-
rus spp.) (Bartos and Gibson 1990), and midges and seed
chalcids (Megastigmus spp.) (Dewey 1989). Siroccocus
strobilinus Preuss is a seed-borne disease that kills white-
bark pine seedlings in niu-series and in natxu-al stands.
Calocypha fulgens (Pers.) Boud. (anamorph = Geniculcden-
dron pyriforme Salt), referred to as a seed or cold fungus,
may cause preemergence seed loss (Hoff and Hagle 1990).
Pre- and postemergence damping-off diseases (Fusarium
spp.) may cause extensive mortality, especially in slowly
emerging seedlings (Landis and others 1990).

179

SEED DISPERSAL

Whitebark pine cones are well suited to the Clark's nut-
cracker, a forest bird that is the major disperser of white-
bark pine seeds (Arno and Hoff 1989; Hutchins and
Lanner 1982; Lanner 1980; Lanner and Vander Wall 1980;
McCaughey and Schmidt 1990; Tomback 1982).

Cone Attributes

Unlike other conifers such as western white pine (Pinus
monticola Dougl.) and limber pine {Pinus flexilis James),
whose cone scales fully open to release seed, whitebark
pine cone scales only open partly. The scale base does not
break away from the cone axis; the wingless seeds are
held firmly in place yet fully exposed and easily accessible
to nutcracker extraction (Eggers 1986; Hutchins and
Lanner 1982). Nutcrackers have strong, pointed bills like
other species of the family Corvidae. Upper portions of
the cone scales are easily broken by nutcrackers along
a thin fracture zone where the massive apophysis tapers
to a thin cross section beneath the seed-bearing cavities
(Lanner 1982). A nutcracker can store over 100 white-
bark pine seeds in its sublingual pouch, a saclike modifi-
cation on the floor of the mouth (Bock and others 1973).

Nutcracker Caching

Whitebark pine regeneration depends almost entirely
on nutcracker seed selection and caching habits. A nut-
cracker appears to discriminate between good, aborted,
insect-infected, or diseased seeds by rattling each seed
in its bill before depositing it in the sublingual pouch
(Hutchins and Lanner 1982; Tomback 1978; Vander Wall
and Balda 1977). Whitebark pine seeds are dispersed by
nutcrackers up to 22 km from their source. Caches are
buried 2 to 4 cm deep with 1 to 25 or more seeds per cache
(fig. 1) (Tomback 1978; Vander Wall and Balda 1977).
One nutcracker can store an estimated 22,000 to 98,000
whitebark pine seeds each year when seed is available
(Hutchins and Lanner 1982; Tomback 1978; Vander Wall
and Balda 1977). Estimates of food requirements indicate
that nutcrackers may store three to five times as many
seeds as needed (Tomback 1983).

Secondary Dispersers

Nutcrackers compete with other seed consumers for
whitebark pine seeds. Red squirrels harvest cones and
seeds from mid-July to early November, greatly reducing
the availability of seed for nutcrackers (Eggers 1986;
Hutchins and Lanner 1982; Reinhart and Mattson 1990;
Smith 1968). Squirrels store cones in middens and seeds
in caches on the forest floor (Hutchins and Lanner 1982;
Reinhart and Mattson 1990).

Grizzlies obtain whitebark pine seeds primarily from
squirrel middens (Kendall 1983). Germination probabil-
ity from squirrel-cached cones and seed is low due to their
deep caching habits (>7 cm) and small numbers of midden
sites. H5^ocotyl growth of whitebark pine is only 3 to 4 cm;
emergence of cotyledons above the soil surface in a

Figure 1 — Nutcracker cache site with five white-
bark pine germinants growing on a litter seedbed
near a log.

squirrel midden is unlikely (Lanner 1982). Squirrels
feed on the cones and seed during the winter and spring
months thus constantly disturbing the middens (Lanner
1982).

Many other animals feed on whitebark pine seed and are
considered secondary dispersers because of the low prob-
ability of germination following their feeding and caching
(Hutchins and Lanner 1982; McCaughey and Schmidt
1990). Other common animals that harvest whitebark
pine seed either from the cones directly or indirectly from
the ground or other animal caches are: birds — ^William's
sapsucker (Sphyrapicus thyroideus), hairy woodpecker
{Picoides villosus), white-headed woodpecker {P. albolar-
vatus), mountain chickadee (Parus gambeli), white-breasted
nuthatch {Sitta carolinensis), Cassin's finch {Carpodacus
cassinii), red crossbill {Loxia curvirostra), pine grosbeak
(Pinicola enucleator), Steller's jay {Cyanocitta stelleri),
raven (Corvus corax), and red-breasted nuthatch {Sitta
canadensis). Mammals — chipmimks {Eutamias spp.),
deer mouse {Peromyscus maniculatus), golden-mantled
groimd squirrel {Spermophilus lateralis), southern red-
backed vole {Clethrionomys gapperi), chickaree {Tamia-
sciurus douglasi) (Hutchins and Lanner 1982; McCaughey
and Schmidt 1990), and black bears {Ursus americanus)
(Craighead and others 1982; Kendall 1983). Secondary
dispersers not only reduce the seed crop but limit the
availability of seed to nutcrackers and thus the probabil-
ity of regeneration.

Usually whitebark pine cone crops are depleted by
cone and seed consumers. Hutchins and Lanner (1982)
described a 3-year period when no whitebark pine cones
fell to the ground other than by animal clipping. In con-
trast, so many cones were produced in 1989 that many
were not utilized by seed consumers and fell to the groimd
(McCaughey and others 1990). Grizzly bears were ob-
served foraging on these cones in 1990 (Reinhart 1990).
The 1989 mast year may have reached the upper limit of

180

cone production, a level that rarely occiirs. Wind storms
may have dislodged the whitebark cones making them
available to ground foraging animals.

The probability of germination is low from seeds that
have fallen to the ground in cones or been dropped by ani-
mals. All whitebark pine seeds sown on the ground are
eaten or taken by seed consxmiers when not protected
with exclosures (McCaughey 1990). This situation oc-
curred in two successive seasons when seeds were sown
during field germination tests. Eight species of rodents
were trapped on site dining the field germination test.
The two most abundant species were the deer mouse and
the southern red-backed vole representing 54 and 23 per-
cent of the total popidation trapped (McCaughey 1990).

GERMINATION

Information on seed storage and laboratory germina-
tion of whitebark pine seeds, under a variety of seed treat-
ment methods, has been summarized and discussed by
several authors in the symposium proceedings "Whitebark
Pine Ecosystems: Ecology and Management of a High-
Mountain Resource" (Schmidt and McDonald 1990). The
following sections extract pertinent information from
those various soiirces.

Seed Storage

Whitebark pine seed has been frozen imder environ-
mentally controlled conditions for up to 20 years (Schubert
1954). Seed viability decreased over that time period from
50 percent to 3 percent. Mirov (1946) found that viability
of whitebark pine seed dropped from 24 percent at time
of collection to 17 percent after being frozen for 8 years,
but dropped to 1 percent after 11 years of storage. Viabil-
ity of whitebark pine seed, as related to storage time, is
dependent on seed maturity when harvested, seed han-
dling prior to storage, and methods and length of time
of stratification for germination tests (McCaughey and
Schmidt 1990).

Controlled Germination

ViabiHty of whitebark pine seed, even iinder controlled
conditions, is highly variable, ranging from 0 to 75 per-
cent. One procedure for germinating whitebark pine
seeds is to soak seeds in running tap water for 1 to 2 days
and stratify moist at 1 to 5 °C for 90 to 120 days in plastic
bags (Krugman and Jenkinson 1974). Stratification time
may vary by seed maturity. Jacobs and Weaver (1990a)
found that 1 month of stratification was sufficient to in-
crease germination from 5 percent to about 40 percent;
longer stratification periods (to 8 months) did not improve
germination. The Coevir d'Alene Nursery in Idaho,
U.S.A., used the following procedm-es for germination
tests for western white pine and whitebark pine:

A. Place seed in nylon mesh bags.

B. Soak seed for 48 hours in ninning tap water. Place
nylon mesh bag in plastic bag.

C. Stratify for 100 days at 1 to 2 °C. Within that 100-
day stratification time resoak the seed for 1 hour each
week.

D. After 100 days, remove seed from stratification and
surface dry.

E. Using a vacuum seeder, place 100 seeds on moist
paper towels (Kimpak) in each of four plastic trays.

F. Place trays in germinator and take counts.

Seedling Morphology

Whitebark pine germinants have thicker stems than
those of associated conifers. Stem diameters range from
2 to 4 mm for whitebark and only 1 to 2 mm for lodgepole
pine {Pinus contorta var. latifolia), Engelmann spruce
{Picea engelmannii Parry ex Engelm.), and subalpine fir
iAhies lasiocarpa [Hook.] Nutt.) (McCaughey 1988). Ger-
minants produce five to 12 cotyledons and grow to heights
of 3 to 5 cm in the first growing season (fig. 2) (McCaughey
and Schmidt 1990). Needles vary in length from 7 to 10
cm with stomata positioned on the dorsal and ventral
sides. The hypoderm is weak, usually one cell in width;
the endodermis has an outer wall that is strongly thick-
ened, and there is only one fibrovascular bundle (Harlow
1931).

Whitebark pine trees commonly have multiple stems
from forking of a single stem or the merging of several
seedlings from a nutcracker cache. Forking occurs on a
small percentage of whitebark pine germinants. Nearly
7 percent of first-year germinants have two forks and
17 percent have three or more forks (McCaughey 1988;
Weaver and Jacobs 1990). Genotypic analyses of multi-
stemmed clumps show that 58, 70, and 83 percent, of
three stands sampled in Alberta, Canada, and Wyoming,
U.S.A., had stems of mixed genetic origin while 42, 30,

Figure 2 — Single whitebark pine seedling germin-
ating on a mineral seedbed and showing newly
emerged cotyledons.

181

and 17 had arisen by branching (Furnier and others
1987; Jacobs and Weaver 1990b; Linhart and Tomback
1985). This indicates that trees of mixed genetic origin
had merged from several genetically individual seedlings
from the same cache.

Whitebark pine germination and root growth are affected
by early season soil temperatures. Seeds germinate when
soil temperatures are between 10 and 40 °C with an opti-
mum temperature range of 25 to 35 °C (Jacobs 1989;
Jacobs and Weaver 1990a). Root growth occurs between
10 and 45 °C with the optimum temperature range being
25 to 35 °C. This optimum range provides conditions
where whitebark roots grow 5 to 15 mm per day (Jacobs
and Weaver 1990a). First-year root growth ranges be-
tween 5 and 18 cm for nutcracker-cached seedlings grow-
ing in a forest environment (McCaughey 1988).

Period of Germination

Germination from a single sowing or caching of white-
bark pine seeds can continue throughout the growing sea-
son and during the next 2 to 3 or more years (McCaughey,
in press b). Because nutcrackers bury seed 2 to 4 cm
deep, seeds may germinate but not emerge above the soil
surface for several days after germination or they may not
emerge at all. The term germinant in this paper refers
to an emergent, since it is difficult to discern if a seed ger-
minated yet died prior to emergence.

Germination Sites

Whitebark pine germinates on a variety of seedbeds,
almost entirely dependent on the choice of the nutcracker.
Whitebark pine grows within 38 habitat-phase combina-
tions in eastern Idaho and western Wyoming (Steele and
others 1983) and 44 in Montana (Pfister and others 1977).
Typical sites for Clark's nutcracker caches are well-drained
and moist substrates, bare soil, forest litter, gravel, rubble,
in small plants and logs, in cracks and fissures on exposed
rock, and in pumice soils (Hutchins and Lanner 1982;
Tomback 1978, 1982; Tomback and others 1990). Regen-
eration is commonly found on burned litter seedbeds fol-
lowing natural and prescribed burns (Morgan and Bimting
1990; Tomback 1986),

Nutcrackers cache whitebark seeds on all aspects, but
the majority of cache sites occur on southeast, south,
southwest, and west-southwest aspects (Tomback and
others 1990). Seed storage in south-aspect windblown
sites probably ensures that some caches are snowfree in
winter and spring so the nutcracker can retrieve them.
Even though fewer seeds are cached on the north aspects,
regeneration densities are highest on north aspects. This
may be due to less successful retrieval by the nutcracker,
or it may be a result of favorable environmental condi-
tions such as moisture and insolation protection during
the germination period (Tomback and others 1990). The
tallest and best formed whitebark pine trees are often
found in high basins or on gentle north slopes (Arno and

Hoff 1989). Drought and insolation mortality of seedlings
on south slopes may contribute to the disproportionate
success of whitebark regeneration on north aspects.

First-Year Germination

Whitebark pine seed germinating from caches by the
nutcracker begins about mid-June and continues through
early September. The number of whitebark pine seeds
germinating in the first year after caching varies by year,
probably in response to timing of spring snowmelt and
early summer rains (fig. 3) (McCaughey 1990). For ex-
ample, in 1988 only 11.5 percent of buried seeds germi-
nated, probably due to below-average precipitation (82
percent of normal). In 1989 precipitation was near normal
and germination of 1988 sown seed was nearly 34 percent
(table 1) (McCaughey 1990).

First-year germination of whitebark pine is significant-
ly affected by percent shade cover (0, 25, and 50 percent),
seedbed condition (mineral, litter, and burned), and sow-
ing depth (surface sown and buried) (table 1) in controlled
field studies in southwestern Montana, U.S.A. (McCaughey
1990). Percent germination of whitebark pine was signifi-
cantly ip = 0.008) higher (20.4 percent) under a 50 percent
shade cover than with no shade (16.7 percent) (table 1)
(McCaughey 1990). Germination of whitebark pine seeds
sown in 1988 was 26 percent on mineral seedbeds and sig-
nificantly less on litter and burned seedbeds, 14.5 and
15.3 percent, respectively (McCaughey 1990). Buried
whitebark pine seeds (2 to 4 cm deep) had significantly
higher germination in 1988 and 1989 (11.5 and 33.7 per-
cent, respectively) than surface-sown seeds that were pro-
tected from seed consumers in the same years (1.8 and
11.5 percent, respectively) (McCaughey 1990).

50-

? 40-

UJ
(D

H 30"
•z.

lU
Q.

lU
>

20-

^ 10-

3
O

Burled, germinating under moist conditions - 1989

JUNE

n- 2.700 seeds

Buried, germinating under dry conditions - 1988
n- 1,440 seeds

JULY

AUGUST ' SEPTEMBER ' OCTOBER

Figure 3 — Cumulative percent germination of white-
bark pine germinants from buried and protected
seeds, sown in 1987 and 1988 respectively, germi-
nating under dry (1988) and moist (1989) conditions.
Data collected from a clearcut at 2,530 m elevation
in southwestern Montana, U.S.A.

182

Table 1 — Percent first-year germination of 1987 and 1988 sown

seeds of whitebark pine as affected by shade cover, seed-
bed condition, and sowing depth. All treatments excluded
seed consumers allowing for surface germination. Data
collected from a clearcut at 2,530 m elevation in south-
western Montana, U.S.A.

First-year germination

Factor

1988

1989

Factor

level

mean

- Percent

Shade cover

^5.2 (a)

16.7 (c)

(percent)

7.9 (a)

18.8 (cd)

6.8 (a)

20.4 (d)

Seedbed

Mineral

8.2 (a)

26.0 (c)

condition

Litter

5.1 (b)

14.5 (d)

Burned

15.3 (d)

Sowing depth

Surface

1.8 (a)

3.5 (c)

Buried

11.5 (b)

33.7 (d)

'Similar and dissimilar letters in parentheses within a column for a factor
represent nonsignificant and significant differences respectively.

Delayed Germination

Delayed germinants are whitebark pine seeds that ger-
minate two or more seasons after being sown or cached.
The European stone pine (Pinus cembra) germinates in
the first, second, and third year after caching (Krugman
and Jenkinson 1974). Whitebark pine seed germinates
at least 3 years following sowing (McCaughey 1990;
McCaughey, in press b; McCaughey and Schmidt 1990).
Although there were absolute value differences, the effects
of shade cover and seedbed condition became nonsignifi-
cant as germination occurred over a 3-year period in a
study of whitebark germination in western Montana,
U.S.A. (table 2) (McCaughey, in press b). Germination
from buried seed (56 percent) remained significantly higher
than for surface-sown seed (7 percent) 3 years after sow-
ing (McCavighey, in press b).

Dormancy of whitebark pine seed is caused by embryo
underdevelopment, physiological embryo dormancy, im-
perviousness of the seed coat and female gametophyte
tissue to oxygen and water uptake, and possibly by depo-
sition of growth inhibitors to the embryo by female game-
tophyte tissue (Pitel 1981; Pitel and Wang 1980, 1990).
Seeds of whitebark pine must mature in a short growing
season in high-elevation forests, and if climatic conditions
slow the maturation process nutcrackers will harvest im-
mature seed. Nutcrackers are normally not concerned
with seed maturity and will also harvest seeds early in
the season before they are mature. Embryo vmderdevel-
opment can be overcome by exposing imbibed seeds to
20 °C for 30 to 60 days (Leadem 1985). Clipping the seed-
coat is a method to overcome physiological barriers to ger-
mination (Pitel 1981).

Under natural conditions, the greatest germination
from cached whitebark pine seeds occurs in the second

year following caching due to delayed dormancy mecha-
nisms (McCaughey 1990, 1992; Tomback 1992). For
example, germination of buried whitebark pine seed in-
creased from 11 percent in the first year following sowing
to 45 percent in the second year. Germination declined to
11 percent in the third year foUovnng sowing (McCaughey,
in press b).

Limiting Factors

Germination rates for whitebark pine are highest from
mid-June through the end of July (McCaughey 1990).
Mortality of first-year seedlings follows the same pattern;
the highest mortality rates occvir when germination is
highest (fig. 4) (McCaughey 1990). Germination exceeds
mortality up to the first of August, resulting in an accu-
mulation of surviving seedlings; however, germination
and mortality rates decrease after the first of August.
Germination continues from early August until the first
of September with total nimabers of survivors remaining
constant because mortality equals germination (McCaughey
1990). Late germinants have the same probability of svir-
vival as early germinants. Factors limiting seedling sur-
vival include:

Microsite and Biotic — Three major factors are:
(1) insolation (heat scorching of seedling stem at ground
surface), (2) drought (drying out of seedling), and (3) ani-
mals (burial, uprooting, or nipping), specifically pocket
gophers (Thomomys talpoides) (McCaughey and Schmidt
1990). Hutchins and Lanner (1982) observed a chipmunk
uproot and consume a whitebark pine seedling. Insola-
tion mortality of whitebark pine was higher on mineral,
litter, and burned seedbeds on nonshaded plots when
compared to 25 and 50 percent shaded plots in a germina-
tion test in southwestern Montana, U.S.A. (McCaughey
1990).

Blister Rust — The introduced disease, blister rust
{Cronartium ribicola), poses the most serious threat to

Table 2 — Percent germination' of 1 987 sown whitebark pine seed
(buried 2 to 4 cm) as affected by shade cover and seed-
bed condition for each of the first 3 years following sowing
and the 3-year total. Data collected from a clearcut at
2,530 m elevation in southwestern Montana, U.S.A.

Emergence

Factor

Factor
level

First
year

Second
year

Third
year

Three-year
total

Shade cover
(percent)

Seedbed
condition

25
50

Mineral
Litter

10
13
12

13
9

Percent

39
51
46

40
50

5
14
15

7
15

247 (a)
61 (a)
60 (a)

51 (a)
60 (a)

'Percent germination for each year is based on the number of seeds that
had not previously emerged.

^Similar and dissimilar letters in parentheses for a factor represent statisti-
cally nonsignificant and significant differences respectively.

183

10
8

6 -I

1988

INSOLATION
DROUGHT

JUNE

JULY AUGUST SEPTEMBER

1989

■1 INSOLATION

m DROUGHT

□ ANIMAL

JUNE 1

JULY 1 AUGUST 1 SEPTEMBER

Figure 4 — Percent first-year mortality of whitebark
pine seedlings by causal agent over time for 1 987
and 1 988 sown seeds. Data collected from a clear-
cut at 2,530 m elevation in southwestern Montana,
U.S.A.

The Yellowstone environment is poorly suited to blister
rust infection because it has a cool, dry climate that is
only marginally suitable for blister rust teliospore germi-
nation (Krebill 1971).

Mountain Pine Beetle — Mountain pine beetle is the
most damaging insect in mature stands of whitebark pine
(Arno and Hoff 1989; McCaughey and Schmidt 1990).
Large numbers of whitebark pine are killed by moimtain
pine beetle when epidemics spread into whitebark pine
stands from the lower elevation lodgepole pine zone. These
epidemics, like blister rust, create conditions where suc-
cession progresses toward dominance by shade-tolerant
conifers such as subalpine fir and Engelmann spruce (Arno
1986).

Fire Suppression — Whitebark pine is serai to subal-
pine fir and other conifers, and periodic fires have helped
to perpetuate the pine (Arno 1986). Past fire intervals
in whitebark pine stands ranged from 50 to 300 years or
more depending on site conditions (Arno 1980). Fires oc-
cur in whitebark pine stands only under severe burning
conditions and are tj^ically non-stand replacing. Fires
sweep across large areas of forest reducing competition
from shade-tolerant conifers (Arno 1976). Nearly 90 years
of fire suppression have reduced the annual acreage of
whitebark pine forests being biu'ned, creating conditions
where shade-tolerant species have taken over what were
once serai whitebark pine stands.

Severe fire conditions are being created by increased
mortality and the buildup of fuels in whitebark pine stands
due to mountain pine beetle and blister rust epidemics.
Stand replacement fires will be the result of this unnatu-
ral fuel loading and will eliminate large cone-producing
stands of whitebark pine.

the survival of whitebark pine in parts of its distribu-
tional range. Blister rust was brought to western North
America in 1910 on a boatload of blister rust-infected
eastern white pine (Pinus strobus L.). Whitebark pine is
the most susceptible to blister rust infection of 14 white
pines rated by Bingham (1972). Blister rust causes exten-
sive damage and mortality in whitebark pine stands in
moist mountain regions in northwestern Montana, north-
ern Idaho, and the Washington Cascades in the United
States (Arno 1986). Damage is light where environmental
conditions are somewhat dry and cold, and there is a de-
crease in low-elevation sources of inoculimi (Hoff and
Hagle 1990).

Blister rust-infected trees may live for several years
before mortality occurs; during that time cone production
is reduced due to loss of upper crowns. Mortality from
rust favors succession toward dominance by whitebark
pine's shade-tolerant associates thus reducing site poten-
tial for whitebark pine regeneration and survival (Arno
1986; Arno and Hoff 1989; McCaughey and Schmidt 1990).

The degree of blister rust infection on whitebark pine
decreases southward for all parts of its range (Hoff and
Hagle 1990; Keane and others 1990). A major population
of whitebark pine exists in the Yellowstone ecosystem of
southwestern Montana and northwestern Wyoming, U.S.A.

VEGETATIVE REPRODUCTION

Whitebark pine can reproduce vegetatively through lay-
ering of lower branches along the ground surface. In the
Mission Range of western Montana vegetative reproduc-
tion was observed from layering of shrublike (krummholz)
whitebark pine (Arno 1981; Arno and Hoff 1989). Although
layering is possible, the vast majority of reproduction is
from seeds.

SUMMARY

Regeneration processes are similar for whitebark pine
and the other stone pines of the world. The initiation,
development, and maturation of cones and seeds for white-
bark pine are characteristic of five-needle pines in general
with seasonal timing of these processes varying between
species. The entire process from cone initiation to cone
and seed maturity takes about 24 months. Cone and seed
development are most affected by variations in climatic
conditions during periods of reproductive bud formation,
pollination, and growth and development. Several insects
and diseases reduce survival of cones and seeds of white-
bark pine.

Regeneration of whitebark pine depends almost exclu-
sively on the seed selection, dispersal, and caching habits

184

of the Clark's nutcracker. Other stone pines of the world
have similar mutualistic relationships with bird species
for seed dispersal.

Whitebark pine seeds are cached by Clark's nutcrackers
on a variety of sites ranging from forest litter to cracks
and fissures in rocks. The majority of regeneration occurs
on south or west aspects. Seed storage in south-aspect
windblown sites ensures that some caches are snow free
in winter and spring for nutcracker retrieval. Regenera-
tion densities are highest on north aspects even though
fewer seeds are cached there because of favorable micro-
site conditions such as moistxu'e and insolation protection.

Germination of whitebark pine begins about mid-June
and continues through early September in the first year
following caching. Delayed germination occurs for up
to 3 or 4 years after caching. Total germination over a
3-year period is highest on litter seedbeds that have 50
percent shade cover, while germination is lower on min-
eral and nonshaded seedbeds (McCaughey, in press b).
Dormancy of whitebark pine seed is caused by embryo
underdevelopment, physiological embryo dormancy, and
imperviousness of the seed coat and female gametoph3^e
tissue to oxygen and water uptake.

Factors limiting survival of whitebark pine seedlings
are microsite, biotic, disease, insects, and fire. Insolation
and drought are microsite-associated factors that cause
seedling mortality. Animals such as pocket gophers bury,
uproot, or nip off young seedlings, and chipmunks uproot
and consimie whole seedlings. Deer mice and southern
red-backed voles and other small mammals consume seed
before they germinate. Blister rust causes mortality or
extensive damage of the cone-bearing portions of the tree.
Whitebark pine is killed by the mountain pine beetle
when epidemics spread into whitebark pine stands from
the lower elevation lodgepole pine zone. Fire suppression
creates conditions where shade-tolerant species such as
Engelmann spruce and subalpine fir, normally killed by
light imderburns, take over through succession what were
once serai whitebark pine stands. Severe fire conditions
are being created by the buildup of fuels in whitebark
pine stands due to increased mortality.

We are beginning to piece together the puzzle of the re-
generation process of whitebark pine. Survival of wildlife
species such as the grizzly is threatened due to a reduc-
tion or loss of localized populations of whitebark pine
caused by global climate change, introduced disease, in-
sects, and succession fi-om fire exclusion. Future research
should define microsite and biotic factors influencing ger-
mination and long-term survival of seedlings. Informa-
tion on regeneration processes from other stone pine spe-
cies from around the world is helping identify key areas
for research on whitebark pine. For example, little infor-
mation is available on the elevational distribution of
whitebark and which habitats are best suited for regen-
eration. Cembra pine {Pinus cembra) grows in the Swiss
Alps fi-om valley bottoms to timberline when management
actions reduce intertree competition. Whitebark pine
management wiU improve from increased knowledge of
regeneration processes of all the stone pines.

REFERENCES

Amo, S. F. 1976. The historical role of fire on the Bitter-
root National Forest. Res. Pap. INT-187. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Forest and Range Experiment Station. 30 p.

Arno, S. F. 1980. Forest fire history in the Northern
Rockies. Journal of Forestry. 78(8): 460-465.

Arno, S. F. 1981. [Unpublished observations]. Missoula,
MT: U.S. Department of Agriculture, Forest Service,
Intermoimtain Fire Sciences Laboratory.

Amo, S. F. 1986. Whitebark pine cone crops - a diminish-
ing source of wildlife food? Western Journal of Applied
Forestry. 1(3): 92-94.

Arno, S. F.; Hoff, R. J. 1989. Silvics of whitebark pine
(Pinus albicaulis). Gen. Tech. Rep. INT-253. Ogden,
UT: U.S. Department of Agriculture, Forest Service,
Intermoimtain Research Station, lip.

Bartos, D. L.; Gibson, K. E. 1990. Insects of whitebark
pine with emphasis on mountain pine beetle. In:
Schmidt, Wyman C; McDonald, Kathy J., comps.
Proceedings — symposium on whitebark pine ecosys-
tems: ecology and management of a high-mountain re-
source; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
171-178.

Bingham, R. T. 1972. Taxonomy, crossability, and relative
blister rust resistance of 5-needled white pines. In: Biol-
ogy of rust resistance in forest trees: proceedings of a
NATO-IUFRO advance study institute; 1969 August
17-24; Moscow, ID. Misc. Publ. 1221. Washington, DC:
U.S. Department of Agriculture: 271-280.

Bock, W. J.; Balda, R. P.; Vander Wall, S. B. 1973. Mor-
phology of the sublingual pouch and tongue muscula-
ture in Clark's nutcracker. Auk. 90: 491-519.

Craighead, J. J.; Sumner, J. S.; Scaggs, G. B. 1982. A de-
finitive system for analysis of grizzly bear habitat and
other wilderness resources. Wildlands Inst. Monogr. 1.
Missoula, MT: University of Montana. 279 p.

Dewey, Jerald E. 1989. [Personal communication to Dale
Bartos of Intermountain Research Station, Logan, UT].
March 27. Missoula, MT: U.S. Department of Agricul-
ture, Forest Service, Northern Region.

Edwards, D. G. W. 1990. Cone prediction, collection and
processing. In: Schmidt, Wyman C; McDonald, Kathy J.,
comps. Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-mountain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
78-102.

Eggers, D. E. 1986. Management of whitebark pine as po-
tential grizzly bear habitat. In: Contreras, G. P.; Evans,
K. E., comps. Proceedings — grizzly bear habitat sympo-
sium; 1985 April 30-May 2; Missoula, MT. Gen. Tech.
Rep. INT-207. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
170-175.

185

Furnier, G.; Knowles, P.; Clyde, M.; Dancik, B. 1987. Ef-
fects of avian seed dispersal on the genetic structure
of whitebark pine populations. Evolution. 41: 607-612.

Harlow, W. M. 1931. The identification of the pines of the
United States, native and introduced, by needle struc-
ture. Tech. Publ. 32. Syracuse, NY: New York State
College of Forestry. 21 p.

Hoff, R. J. 1992. Unpublished data on file at: U.S. Depart-
ment of Agriculture, Forest Service, Intermountain Re-
search Station, Forestry Sciences Laboratory, Moscow,
ID.

Hoff, R. J.; Hagle, S. 1990. Diseases of whitebark pine
with special emphasis on white pine blister rust. In:
Schmidt, Wyman C; McDonald, Kathy J., comps.
Proceedings — symposium on whitebark pine ecosys-
tems: ecology and management of a high-mountain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
179-190.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Jacobs, J. S. 1989. Temperature and light effects on seed-
ling performance of Pinus albicaulis. Bozeman, MT:
Montana State University. 39 p. Thesis.

Jacobs, J.; Weaver, T. 1990a. Effects of temperature and
temperature preconditioning on seedling performance
of whitebark pine. In: Schmidt, Wyman C; McDonald,
Kathy J., comps. Proceedings — symposium on white-
bark pine ecosystems: ecology and management of a
high-mountain resource; 1989 March 29-31; Bozeman,
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart-
ment of Agriculture, Forest Service, Intermountain
Research Station: 134-139.

Jacobs, J.; Weaver, T. 1990b. Occurrence of multiple
stems in whitebark pine. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and manage-
ment of a high-moimtain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 156-159.

Keane, R. E.; Arno, S. F.; Brown, J. K.; Tomback, D. F.
1990. Simulating disturbances and conifer succession
in whitebark pine forests. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and manage-
ment of a high-mountain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 274-288.

Kendall, K. C. 1983. Use of pine nuts by grizzly and black
bears in the Yellowstone area. In: International Confer-
ence on Bear Research and Management. Madison, WI:
International Bears Association. 5: 166-173.

Knight, R. R.; Blanchard, B. M.; Mattson, D. J. 1987.
Yellowstone grizzly bear investigation: report of the
interagency study team, 1987. Bozeman, MT: U.S. De-
partment of the Interior, National Park Service. 80 p.

Krebill, R. G. 1971. Conditions for teliospore germination
in Yellowstone National Park environment. Unpub-
lished report on file at: U.S. Department of Agriculture,
Forest Service, Intermountain Research Station,
Ogden, UT.

Krugman, S. L.; Jenkinson, J. L. 1974. Pinus L. pine.
In: Schopmeyer, C. S., ed. Seeds of woody plants in the
United States. Agric. Handb. 450. Washington, DC:
U.S. Department of Agriculture, Forest Service: 598-638.

Landis, T. D.; Tinus, R. W.; McDonald, S. E.; Barnett,
J. P. 1990. The biological component: nursery pests and
mycorrhizae. Vol. 5. The container tree nursery manual.
Agric. Handb. 674. Washington, DC: U.S. Department
of Agriculture, Forest Service. 171 p.

Lanner, R. M. 1980. Avian seed dispersal as a factor in
the ecology and evolution of limber and whitebark
pines. In: Sixth North American forest biology work-
shop proceedings; 1980 August; Edmonton, AB.
Edmonton, AB: University of Alberta: 14-48.

Lanner, R. M. 1982. Adaptations of whitebark pine for
seed dispersal by Clark's nutcracker. Canadian Journal
of Forest Research. 12(2): 391-402.

Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of lim-
ber pine seed by Clark's nutcracker. Journal of For-
estry. 78(10): 637-639.

Leadem, C. L. 1985. Seed dormancy in three Pinus species
of the Inland Mountain West. In: Schmidt, Wyman C,
comp. Proceedings — future forests of the Moimtain
West: a stand culture symposium; 1986 September 29-
October 3; Missoula, MT. Gen. Tech. Rep. INT-243.
Ogden, UT: U.S. Department of Agriculture, Forest
Service, Intermountain Research Station: 117-124.

Linhart, Y.; Tomback, D. F. 1985. Seed dispersal by nut-
crackers causes multi-trunk growth form in pines.
Oecologia. 67: 107-110.

McCaughey, W. W. 1990. Biotic and microsite factors
affecting Pinus albicaulis establishment and survival.
Bozeman, MT: Montana State University. 78 p.
Dissertation.

McCaughey, W. W. 1988. Unpublished data on file at:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station, Forestry Sciences Labora-
tory, Bozeman, MT.

McCaughey, W. W. 1992. Unpublished data on file at:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station, Forestry Sciences Labora-
tory, Bozeman, MT.

McCaughey, W. W. [In press a]. Seasonal maturation of
whitebark pine seed in the Greater Yellowstone ecosys-
tem. In: Schullery, Paul; Despain, Don, comps. Proceed-
ings of the 1st biennial scientific conference on the
Greater Yellowstone Ecosystem — plants and their envi-
ronments; 1991 September 16-17; Mammoth Hot
Springs, Yellowstone National Park, WY. Tech. Rep.
USDI National Park Service.

McCaughey, W. W. [In press b]. Delayed germination and
seedling emergence of Pinus albicaulis in a high eleva-
tion clearcut in Montana, U.S.A. In: Edwards, D. G.,
comp. Proceedings — symposium: seed dormancy and
barriers to germination; 1991 April 23-26; lUFRO

186

project group P2. 04-00 meeting on seed problems;
Victoria, BC. Victoria, BC: Forestry Canada.

McCaughey, W. W.; Amo, S. F.; Tomback, D. F. 1990.
Unpublished personal observations from Henderson
Mountain near Cooke City, MT, U.S.A.. On file at: U.S.
Department of Agriciilture, Forest Service, Intermoun-
tain Research Station, Forestry Sciences Laboratory,
Bozeman, MT.

McCaughey, W. W.; Schmidt, W. C. 1990. Autecology of
whitebark pine. In: Schmidt, Wjonan C; McDonald,
Kathy J., comps. Proceedings — symposium on whitebark
pine ecosystems: ecology and management of a high-
mountain resource; 1989 March 29-31; Bozeman, MT.
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department
of Agriculture, Forest Service, Intermoimtain Research
Station: 85-96.

Mirov, N. T. 1946. Viability of pine seed after prolonged
cold storage. Journal of Forestry. 44(3): 193-195.

Mirov, N. T. 1967. The genus Pinus. New York: The Ronald
Press. 602 p.

Morgan, Penny; Bunting, Stephen C. 1990. Fire effects
in whitebark pine forests. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — symposiimi
on whitebark pine ecosystems: ecology and manage-
ment of a high-mountain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 166-170.

Pitel, J. A. 1981. Improved germination of whitebark
pine {Pinus alhicaulis Engelm.) seeds. Chalk River, ON:
Canadian Forestry Service, Petawawa National For-
estry Institute; unpublished report. 5 p.

Pitel, J. A.; Wang, B. S. P. 1980. A prehminary study of
dormancy in Pinus albicaulis seeds. Bi-Monthly Re-
search Notes. 36(1): 4-5.

Pitel, J. A.; Wang, B. S. P. 1990. Physical and chemical
treatments to improve germination of whitebark pine
seeds. In: Schmidt, Wyman C; McDonald, Kathy J.,
comps. Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-movmtain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agriciil-
ture, Forest Service, Intermoimtain Research Station:
130-133.

Reinhart, D. P. 1990. [Personal commmiication]. August.
Bozeman, MT: U.S. Department of the Interior;
Montana State University, Interagency Grizzly Bear
Project.

Reinhart, D. P.; Mattson, D. J. 1990. Red sqiiirrels in the
whitebark zone. In: Schmidt, Wyman C; McDonald,
Kathy J., comps. Proceedings — symposium on whitebark
pine ecosystems: ecology and management of a high-
moxmtain resource; 1989 March 29-31; Bozeman, MT.
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department
of Agriculture, Forest Service, Intermoimtain Research
Station: 256-263.

Schmidt, W. C; Lotan, J. E. 1980. Phenology of common
forest flora of the Northern Rockies— 1928 to 1937. Res.
Pap. INT-259. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Forest and Range
Experiment Station. 20 p.

Schmidt, W. C; McDonald, K. J., comps. 1990.
Proceedings — symposium on whitebark pine eco-
systems: ecology and management of a high-mountain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station.
386 p.

Schubert, G. H. 1954. Viability of various coniferous seeds
after cold storage. Journal of Forestry. 52: 446-447.

Smith, C. C. 1968. The adaptive nature of social organiza-
tion in the genus of tree squirrels Tamiasciurus . Eco-
logical Monographs. 38: 31-63.

Steele, R.; Cooper, S. V.; Ondov, D. M.; Roberts, D. W.;
Pfister, R. D. 1983. Forest habitat types of eastern
Idaho-western Wyoming. Gen. Tech. Rep. INT-144.
Ogden, UT: U.S. Department of Agriculture, Forest
Service, Intermountain Forest and Range Experiment
Station. 122 p.

Tomback, D. F. 1978. Foraging strategies of Clark's nut-
cracker. Living Bird. 16(1977): 123-160.

Tomback, D. F. 1982. Dispersal of whitebark pine seeds
by Clark's nutcracker: a mutualism hypothesis. Journal
of Animal Ecology. 51(2): 451-467.

Tomback, D. F. 1983. Nutcrackers and pines: coevolution
or coadaptation? In: Nitecki, M. H., ed. Coevolution.
Chicago and London: University of Chicago Press:
179-223.

Tomback, D. F. 1986. Post-fire regeneration of krummholz
whitebark pine: a consequence of nutcracker seed cach-
ing. Madrono. 33(2): 100-110.

Tomback, D. F. 1992. Unpublished data on file at: Depart-
ment of Biology, University of Colorado at Denver,
Denver, CO.

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Co-
evolution of whitebark pine and nutcrackers: implica-
tions for forest regeneration. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — symposiima
on whitebark pine ecosystems: ecology and manage-
ment of a high-mountain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 118-129.

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of
the Clark's nutcracker and the pinon pine for efficient
seed harvest and dispersal. Ecological Monographs.
47: 89-111.

Weaver, T.; Jacobs, J. 1990. Occurrence of multiple stems
in whitebark pine. In: Schmidt, Wjonan C; McDonald,
Kathy J., comps. Proceedings — symposium on white-
bark pine ecosystems: ecology and management of a
high-mountain resource; 1989 March 29-31; Bozeman,
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart-
ment of Agriculture, Forest Service, Intermountain Re-
search Station: 156-159.

187

JAPANESE STONE PINE CONE
PRODUCTION ESTIMATED FROM CONE
SCARS, MOUNT KISOKOMAGATAKE,
CENTRAL JAPANESE ALPS

Ikuko Nakashinden

Abstract — The past cone production of the Japanese stone pine
(Pinus pumila) on Mount Kisokomagatake was reconstructed
from cone scars. The main cone production was by larger indi-
viduals in the lower part of the scrub zone; it was greater than
that in the upper part where growth and crop are controlled
mainly by strong winter winds and snow.

The Japanese stone pine {Pinus pumila Regal) is a
genetic dwarf pine distributed over the eastern part of
Siberia and China, the northern part of Korea, and Japan.
In Japan, most of its habitats are located on high moun-
tains. The climax scrub zone is formed above the timber-
line of subalpine coniferous forests. At the southern mar-
gin of its range (Chubu District) the scrub zone occupies
the altitudinal zone from 1,961 to 3,192 m (Yanagimachi
and Ohmori 1991).

The Japanese stone pine propagates in two different
ways: one by the expansion of shoots and the other by
seeds. The latter is the most important for spatial ex-
pansion (Okitsu 1990). Seeds are dispersed by seed-
storing corvids {Nucifraga caryocatactes) (Hayashida
1989; Tomback 1990). Studies concerning pine cone
production have begun recently. Okitsu and Mizoguchi
(1990, 1991) evaluated pine cone productivity based on
the individual growth sizes and scrub sizes. In addition,
studies of the relationship between environment of
the habitats and cone production have commenced
(Nakashinden 1990). These studies are limited to 1 year's
cone production data. It is notable that past cone produc-
tion can be estimated from cone scars. This was done on
whitebark pine {Pinus albicaulis) in the United States
(Morgan and Bimting 1992; Weaver and Forcella 1986).
Saito and others (1989) reported that cone scars are also
recognizable in the Japanese stone pine in Hokkaido, but
little information about it was given in their report.

This author developed the method to reveal the yearly
change in cone production for the Japanese stone pine and
succeeded in reconstructing pine cone production over the
past 15 years (Nakashinden 1991). Using this cone scar
method, chronological pine cone production of the Japa-
nese stone pine is described and the relationship to the
environment of its habitats is examined.

STUDY AREA AND METHODS

The study area is located on Moimt Kisokomagatake,
whose simimit is 2,956 m, in the Central Japanese Alps,
in Chubu District (fig. 1). This location is one of the
southernmost parts of the Japeinese stone pine
distribution.

The habitats examined are located on the mountain
slopes between 2,600 and 2,950 m elevation, including
the summits of Mount Kisokomagatake and Mount
Chausudake (2,652 m). In this altitudinal zone, the
Japanese stone pine is dominant, forming its own vege-
tation zone above the subalpine forest of Abies veitchii,
Tsuga diversifolia, Betula ermanii, and Alnus
maximowiczii.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Ikuko Nakashinden is Graduate Student, Department of Geography,
University of Tokyo, Bunkyo-ku, Tokyo 113, Japan.

Figure 1— Location of Mount Kisokomagatake.

188

Pine Cone Scars

The positions of pine cones of the Japanese stone pine
are restricted to the first nodes of the annual shoots.
They matiire in the second year after fruiting. In
September, nutcrackers harvest cones of the Japanese
stone pine as a food source. When the pine cone is picked,
a scar of the cone remains on the stem (Nakashinden
1991; Saito and others 1989). Each scar represents one
cone. From these scars, the number of pine cones can be
calciilated. The production can be determined from the
location of the nodes on the stems. This cone scar method
was used on whitebark pine in the United States, and the
scars could be measured for 6 to 8 years (Weaver and
Forcella 1986), or 6 to 12 years (Morgan and Bunting
1992). For the Japanese stone pine, the scars could be
counted for 15 to 20 years (Nakashinden 1991).

Using this method, 32 plots (1 m^) at different altitudes
from 2,600 m near the forest hne to 2,950 m near the
summit were investigated. Ten stems in each plot were
selected for counting pine cone scars and for the measure-
ment of individual stem growth for the 15 years preceding
1990.

Distribution, Geomorphology, and
Snow

Using color aerial photographs and field observation,
the Japanese stone pine scrubs in the study area were
delineated and mapped. Some topographical maps were
made during the field survey in order to examine the mi-
croscale landform beneath the scrubs. The snow melting
status and snow melting period near the simimit of
Moxmt Kisokomagatake was observed and recorded on
photographs in each spring season from 1989 to 1991.

250 -

200 -

150 -

100 -

50-

r— T

76 77 78 79 80 81 82

Year

II I ' I I I I

83 84 85 86 87 88 89 90

Figure 2 — Annual cone crops of Japanese stone
pine (1976-90). Cone crops were reconstructed
from cone scars on the stems. Total number from
32 plots is shown for each year.

o
o
E

o
O

4.0

3.0

2.0

1.0

1983

J L

82 84

Year

86 88

Figure 3 — Mean annual stem elongation of
Japanese stone pine (1976-90). Ten stems
each in 32 plots were measured.

CONE PRODUCTION OVER 15 YEARS

The past 15 years of pine cone production was recon-
structed on the study area (fig. 2). Mast years were
recognized in 1981, 1984, 1988, and 1989, showing a
3- or 4-year periodicity of cone production. The lowest
production during the past 15 years occurred in 1982 and
1983. In 1983, the annual shoot length was also the low-
est during the 15-year period (fig. 3). The year 1983 is
judged to be one of the worst years for both cone crop and
growth.

On the individual plots, both mast and fail years oc-
curred regardless of their environmental differences.
For the mast years, 68 percent of the 32 plots showed
high production in 1981, and in 1984 it was 65 percent.
For the fail years, 59 percent of the plots failed in cone
production in 1982, and in 1983 it was 63 percent. These
occurrences were dependent on the climate and not on
the growth forms or the location of scrubs.

DISTRffiUTION PATTERNS BY
ALTITUDE

On Mount Kisokomagatake, the Japanese stone pine
scrubs dominate and form a vertical zone above the subal-
pine forest at an elevation between 2,600 and 2,950 m.
However, the distribution patterns were characterized by
the altitude. On the upper part from 2,850 up to 2,950 m,
near the summit of Moimt Kisokomagatake, the distribu-
tion pattern of the scrubs shows a stripe pattern (fig. 4).

The scrubs are fragmented by huge granite rocks, tors,
stone bank terraces, trails, and alpine plant communities
such as Arctous alpinus vai.Japonicus, Diapensia lap-
ponia var. obovata, Oxytropis japonica Maxim, Leonto-
podium shinanense Kitam, Gentiana algida Pall., and
Loiseleuria procumhens Desv. The lower part between
2,850 and 2,600 m, near the boimdary to the subalpine
forest shows an extensively spreading scrub, covering the
slopes entirely from the ridges down to the forest limit
(fig. 5).

189

Figure 4 — Distribution pattern of Japanese stone
pine scrubs in tine upper part of the Mount
Kisokomagatake study area. Distribution is frag-
mented by huge rocks, tors, stone bank terraces,
trails, and alpine plant communities.

GROWTH FORM AND CONE
PRODUCTION RELATED TO
ALTITUDE

Growth form and pine cone production are ditTerent be-
tween the upper and the lower part of the Japanese stone
pine scrub zone (table 1). The mean scrub height, stem
length, and diameter at the stem base in the lower part
are almost twice as large as those in the upper part. The
mean pine cone production per stem is 1.5 times more
than in the upper part. Although the niunber of stems
per square meter in the upper part was twice that of the
lower part, the Japanese stone pine scrubs in the lower
part were larger in size and produced more cones than the
scrubs in the upper part.

DISCUSSION AND CONCLUSIONS

On Moimt Kisokomagatake, the mean summer tem-
perature (July and August) was 10 to 12 °C at 2,600 m
(Koizumi 1974) and 10 °C (Nakashinden 1990, unpub-
lished) at 2,850 m. Concerning the lower parts of the
Japanese stone pine scrub zone, the thermal conditions
indicate that the Japanese stone pine scrub zone is in-
cluded in the thermal subalpine zone (Yanagimachi and
Ohomori 1991). There, the Japanese stone pine scrubs
are considered to grow and produce cones at the same
rates as in the subalpine zone.

The scrubs in the upper part are explained by other en-
vironmental factors unfavorable for growth and cone pro-
duction. In the Japanese stone pine scrub zone, the scrub
height is related to winter wind strength and snow depth
of their habitats (Okitsu 1987). In high-moimtain areas,
the Japanese stone pine needs to be protected by snow-
pack to evade physical damages or dehydration caused by
strong winds. Short scrub height indicates strong winds
and shallow snow cover at the habitats during the winter
season, while tall scrub height indicates moderate winds
and heavy snowpack.

On Movmt Kisokomagatake, the depth of snow varies
from 10 cm to 3 m according to landforms and wind veloc-
ity (Koiziuni 1974). Especially around the simimit, most
of the snow is blown away from the windward slope by
strong westerly winds (fig. 6). Snow accumulates on the
leeward side of the ridges or rocks. On the windward
slope, characterized by shallow snow cover, the Japanese
stone pine grow close to the ground, thus protecting them-
selves from strong winds (fig. 7). Therefore, vertical
growth of the Japanese stone pine scrubs is disturbed by
winter winds.

Variation in snow depth caused differences in the
length of the snow melting period. According to 3 years'
observation by this author, snow on the Japanese stone
pine scrubs located on windward slopes started to melt

Figure 5 — Distribution pattern of Japanese
stone pine scrubs in the lower part of the
Mount Kisokomagatake study area. They
spread extensively and cover the slopes en-
tirely from the ridges down to the forest line.

Table 1 — The growth and pine cone production in each part of
the distribution zone viewed from different altitude of
quadrats. The boundary of the upper and the lower part
is 2,850 m

Cone production and growth

Upper part

Lower part

Cone production (pieces)

4.1

6.4

Mean scrub height (cm)

36.2

89.2

Mean stem length (cm)

92.0

180.0

Mean diameter at stem base (cm)

2.2

4.2

Mean annual stem elongation (cm)

2.4

3.0

Number of stems/m^

38.5

20.0

'The values indicate the means of 20 plots for the upper part and 12 plots
for the lower part.

190

Figure 6 — Mount Kisokomagatake summit in
early spring (1 991 ). Snow depth varies accord-
ing to wind velocities and topography. The
western-most part (left) where snow cover was
thin due to strong westerly winds shows earliest
snow melt in spring.

at the end of April. On the leev^^ard slopes, the scrubs
became completely free of snow by the middle of June.
On the slopes where snow remains imtil the end of July,
the Japanese stone pine is replaced by alpine snowbed
vegetation. On the other hand, shallow snow cover causes
early snow melt and solifluction is very common. Solifluc-
tion forms stone banked terraces and cuts off the roots of
young plants. For this reason, the Japanese stone pine
scrubs in the upper part are not able to cover all the
slopes and thus show a fragmented distribution pattern.

These severe environmental conditions are not favor-
able for growth or cone production of the Japanese stone
pine. The main cone production is performed by large and

Figure 7 — Creeping Japanese stone pine in the
upper part of the study area. The scrubs' height is
10-15 cm. The stem has branches flagged toward
the leeward side. The plot shows one of the low-
est cone productions (only one cone for 1 0 stems)
during these 1 5 years.

tall scrubs at the lower elevations where winter winds are
moderate and snow is evenly distributed.

ACKNOWLEDGMENTS

I thank Professor Dr. F.-K. Holtmeier, Dr. Wyman C.
Schmidt, and Dr. W. Schonenberger for their invitation
to attend the workshop. I thank Dr. Professor Hiroo
Ohmori and Dr. Takeei Koizumi for their valuable advice
in carrying out this study. My thanks also go to Dr. Tad
Weaver and Dr. Penelope Morgan for providing me with
information about cone scar methods on whitebark pine.
Dr. Mitsuhiro Hayashida offered me valuable information
in preparation for the workshop. I am grateful to all my
colleagues in the Geography Department of Tokyo
Gakugei University, who enthusiastically helped me in
my field research on Mount Kisokomagatake. I thank
Mr. Wataru, Mrs. Joyce Mukai, and Mr. Garrett Vance
who checked my manuscript. And I thank Dr. Miki
Kobayashi who always encourages me to keep studying
Pinus pumila.

REFERENCES

Hayashida, M. 1989. Seed dispersal and regeneration pat-
terns of Pinus parviflora var. pentaphylla on Mt. Apoi
in Hokkaido. Research Bulletins of the College Experi-
mental Forests, Faculty of Agriculture, Hokkaido Uni-
versity. 46-1: 177-190.

Kajimoto, T. 1989. Above-groimd biomass and litterfall of
Pinus pumila scrubs growing on the Kiso Mountain
Range in central Japan. Ecological Research. 4: 55-69.

Koiziuni, T. 1974. Landschaftsokologische Unter-
suchungen in der Alpine Stufe des Kisokomagatake in
der Japanischen Zentralalpen mit Besonderer Beriick-
sichtigung der Vegeta Strukturbodens. Japanese Jour-
nal of Ecology. 24(2): 78-91.

Mizoguchi, T.; Okitsu, S. 1987. Cone production of Pinus
pumila Regal in high mountains in central Japan with
special attention to basal diameter and annual elonga-
tion on it. In: Transactions of the 98th annual meeting
of the Japanese Forestry Society; 1987 April; K5aishu:
363-364.

Morgan, P.; Bunting, S. C. 1992. Technical note: cone
crops of whitebark pine. Western Journal of Applied
Forestry. 7(3): 71-73.

Nakashinden, I. 1990. Pine cone production of Pinus
pumila and their environment in Mt. Kisokomagatake,
Nagano prefecture. In: Transactions of the 38th autum-
nal meeting of the Association of Japanese Geogra-
phers; 1990 September; Niigata: 104-105.

Nakashinden, I. 1991. Pine cone production of Pinus
pumila reconstructed from pine cone scars. In: Transac-
tions of the 39th spring meeting of the Association of
Japanese Geographers; 1991 April; Tokyo: 28-29.

Okitsu, S. 1984b. Comparative studies on the Japanese
alpine zone, with special reference to the ecology of
Pinus pumila thickets. Geographical Review of Japan.
57(11): 170-172.

Okitsu, S. 1987. Pinus pumila zone. In: Ito, K., comp. Veg-
etation of Hokkaido. Hokkaido, Japan: Hokkaido Univer-
sity Press: 129-167.

191

Okitsu, S.; Ito, K. 1983. Dynamic ecology of the Pinus
pumila community of Mts. Taisetsu, Hokkaido, Japan.
The Environmental Science, Hokkaido University. 6(1):
151-184.

Okitsu, S.; Mizoguchi, T. 1986. Cone production of Pinus
pumila community on Mt. Senjo, Akaishi Range, cen-
tral Japan. In: Transactions of the 97th annual meeting
of the Japanese Forestry Society; 1986 April;
Utsunomiya: 343-344.

Okitsu, S.; Mizoguchi T. 1990. Relation between cone
production and stem diameter and stem elongation
of Pinus pumila of Japanese high mountains. Japanese
Journal of Ecology. 40(2): 49-55.

Okitsu, S.; Mizoguchi, T. 1991. Cone production of Pinus
pumila scrubs. Japanese Journal of Ecology. 41(2):
101-107.

Saito, S.; Kawabe, M.; Nakaoka, T. 1989. On the forest
vegetation of Mt. Upepesanke, Tokachi, Hokkaido. (1):
On a thicket of Pinus pumila at the ridge of 1610 m.

Bulletin of the Higashi Taisetsu Museum of Natural
History. 11: 23-34.
Saito, S.; Kawabe, M. 1990. On the forest vegetation of
Mt. Higashi-Nupukaushinupuri, Tokachi, Hokkaido.
(2): On two thickets of Pinus pumila. Bulletin of the
Higashi Taisetsu Museimi of Natural History. 12:
17-29.

Tomback, D.; Linhart, Y. 1990. The evolution of bird-
dispersed pines. Evolutionary Ecology. 4: 185-219.

Weaver, T.; Forcella, F. 1986. Cone production in
Pinus albicaulis forests. In: Shearer, R. C, comp.
Proceedings — conifer tree seed in the Inland Mountain
West; 1985 August 5-6; Missoula, MT. Gen. Tech. Rep.
INT-203. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermountain Research Station: 68-76.

Yanagimachi, O.; Ohmori, H. 1991. Ecological status of
Pinus pumila scrub and the lower boimdary of the
Japanese alpine zone. Arctic and Alpine Research.
23(4): 242-435.

192

EFFECTS OF SEED DISPERSAL BY
CLARK'S NUTCRACKER ON EARLY
POSTFIRE REGENERATION OF
WHITEBARK PINE

Diana F. Tomback

Abstract — Clark's nutcrackers {Nucifraga columbiana) bury
whitebark pine (Pinus alhicaulis) seeds in recently burned ter-
rain. Studies of early postfire whitebark pine regeneration in the
Northern Rocky Mountains indicate that seed dispersal by nut-
crackers (1) may be more effective than seed dispersal by wind in
restocking large bums, particularly when the seed source is unfa-
vorably positioned with respect to prevailing winds, and (2) pro-
duces a negative exponential distribution when whitebark pine
regeneration density is plotted vs. distance from the seed source.
Whitebark pine seeds can remain dormant for one or more years
after dispersal, which results in continuity of regeneration over
time.

1986), although updrafts and storms may disperse some
seeds over longer distances. Often, the regeneration from
nutcracker caches is in the form of whitebark pine seed-
Ung clusters (for example see Tomback 1982, 1986; Tomback
and others 1990, 1993), which may result in mature multi-
genet tree clusters (see Tomback and Schuster, these pro-
ceedings, and references therein). In this paper I summa-
rize the results of recent studies that compare patterns of
early postfire regeneration of the bird-dispersed white-
bark pine and its wind-dispersed forest associates.

REGENERATION IN WESTERN
MONTANA

Studies of two large, severe subalpine bums in western
Montana (Bitterroot National Forest, Ravalli County)
illustrate how seed dispersal by nutcrackers affects pat-
terns of early postfire regeneration of whitebark pine and
associated conifers. These two study areas were selected
for similarities in topography, severity of bum, forest
types, and time elapsed since fire. In this region of west-
em Montana, whitebark pine is abundant between about
2,290 and 2,620 m elevation (Amo and Hoff 1989; Pfister
and others 1977), and the prevailing winds in late sum-
mer and fall are from the west and southwest (Finklin
1983). The prefire forest in both study areas was a serai
whitebark pine community. Reported results fi-om these
studies are summarized here from Tomback and others
(1990, 1993).

The Sleeping Child Bum resulted from a wildfire in
1961 that bumed about 11,350 ha of forest on the west
slope of the Sapphire Range. The nearest whitebark pine
seed sovu*ce is mature forest above 2,250 m on the eastern
edge of the bum. Subalpine fir, Engelmann spmce (Picea
engelmannii), and lodgepole pine (Pinus contorta) are also
present in this upper seed source as well as around the
lower perimeters of the bum. Because the prevailing
winds are fi*om the west, vdnd-dispersed seeds are irregu-
larly dispersed into the eastem portion of the bum fi'om
the upper seed source.

In 1987, we established a 3.6-km-long transect along
an eastwest trending ridge, running from the seed soiirce
into the northem center of the burn (fig. 2 in Tomback
and others 1990). Elevation along the ridge decreases
from about 2,480 to 2,150 m with increasing distance
from the eastern edge of the bum. Every 150 m along the
transect we established parallel quadrats, each 50 m long,
on the ridgetop, north, and south aspects of the ridge for

Fire is an ecologically important, recurring event in
most western North American montane forests, including
those regions where whitebark pine {Pinus albicaulis) is
found. Fire-prone areas are those with fuel accumulation,
dry, windy weather, and periods of frequent lightning
storms. In the Northem Rocky Mountains, fire burns
whitebark pine communities every 50 to 300 years (Amo
1980; Amo and Petersen 1983; Romme 1980). Stand-
replacing fires result in the renewal of successional white-
bark pine communities, and underbums kill the less fire-
resistant subalpine fir {Abies lasiocarpa). In the absence
of fire in these commimities, whitebark pine is often re-
placed by more shade-tolerant conifers, particularly sub-
alpine fir (Amo 1986; Amo and Hoff 1989).

Studies of early postfire regeneration, defined here as
the first 3 decades after fire, provide information on the
dynamics of seed dispersal and seedling establishment for
whitebark pine and associated conifers. Clark's nutcrack-
ers {Nucifraga columbiana) are the principal dispersal
agents for whitebark pine seeds, in contrast to other coni-
fers whose seeds are wind-dispersed. Nutcrackers bury
caches of one to 15 or more ripe whitebark pine seeds at
distances from parent trees of a few meters to 22 km
(Hutchins and Lanner 1982; Tomback 1978, 1982; Vander
Wall and Balda 1977; see review in Tomback and Linhart
1990). In contrast, most vraid-dispersed seeds land only
within about 120 m of parent trees (McCaughey and others

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Diana F. Tomback is Associate Professor of Ecology, Department of
Biology and Center for Environmental Sciences, University of Colorado
at Denver, Denver, CO 80217-3364.

193

a total of 63 quadrats. Quadrat widths varied in preset
increments from 1.25 to 15 m, with respect to local densi-
ties of whitebark pine. For each quadrat we recorded all
conifer regeneration.

The Saddle Moimtain Burn resulted from a wildfire in
1960 that spread over 1,240 ha of forest in the Bitterroot
Mountains west of Lost Trail Pass. Running northeast
to southwest in the longest dimension, the burned area
ranges in elevation from about 1,950 to 2,475 m. On the
southwest edge of the burn above 2,275 m, the whitebark
pine seed source is positioned so that the burned area is
leeward with respect to prevailing winds. Subalpine fir,
Engelmann spruce, and lodgepole pine also occur in this
upper seed source and, with Douglas-fir {Pseudotsuga
menziesii), around the lower perimeters of the burn.

In 1988 we established two parallel 3.2-km-long tran-
sects beginning at the seed source and heading northeast
through the southern portion of the burn (fig. 4 in Tomback
and others 1990). We set up 41 quadrats and recorded all
regeneration according to the general methodology out-
lined earlier for the Sleeping Child Burn (for details of
methodology, see Tomback and others 1990, 1993).

Mean whitebark pine regeneration density sampled per
quadrat in the Sleeping Child Burn was much greater
than that of its shade-tolerant competitors, subalpine fir

Table 1 — Mean density and standard deviation (S.D.) per quadrat
for conifer^ regeneration 26 years and 28 years after fire
in tlie Sleeping Child and Saddle Mountain Burns,
respectively. Based on Tomback and others (1 990)

Regeneration sites per
ABLA PIEN PICO PSME

PIAL

Sleeping Child

X 0.070 0.008 0.004 0.039

S.D. .104 .014 .009 .066

Saddle Mountain

X .044 .046 .030 .323

S.D. .050 .068 .044 .355

0.019
.032

TIAL = whitebark pine, ABLA = subalpine fir, PIEN
PSME = Douglas-fir.

Engelmann spruce.

and Engelmann spruce, and greater than the shade-
intolerant lodgepole pine as well (table 1). Scatterplots of
quadrat densities vs. distance from the upper seed source
revealed interesting differences in regeneration patterns
among the conifers (fig. 1). The scatterplot for whitebark
pine was a negative exponential curve, with the highest
densities near the seed source and a long tail indicating

0.512

0.400 -

0.200 -

'3)
c

p. albicaulis

• • • • •
• • • • •

.• • . J J I It •• • •

0.296 |- p. contorta

0.200

0.100

• • • •

- • • • •

• • •

0.08 r A. lasiocarpa

0.05

0.02

0.056 r p engelmannii

0.040

• • •

0.020

• I

• I •
• • •

I I »» w I l,« « i »» . I » y I « I I

1000

2000

3000 3650

1000

2000

3000 3650

Distance (m)

Figure 1 — Scatterplots of regeneration density per quadrat versus distance from the whitebark pine
seed source for four conifer species in the Sleeping Child Burn. Clockwise from left: whitebark pine,
lodgepole pine, Engelmann spruce, subalpine fir. From Tomback and others (1990).

194

some regeneration beyond 3.6 km. Quadrat density vs.
distance in subalpine fir showed a negative linear rela-
tionship, indicating that the primary seed soiirce for this
species was also the upper, unbumed forest. In contrast,
the scatterplot for Engelmann spruce was nearly flat,
with the highest values at the lower end of the transect,
suggesting that seeds may be blown into the ridge area
from both the higher and lower seed sources. For lodge-
pole pine, densities increased with distance from the up-
per, iinbumed forest (are higher at lower elevations).
Before the fire, the lower and middle elevations of this
area were dominated by lodgepole pine, and thus much of
the postfire regeneration may have come from seeds from
serotinous cones within the bum (Lotan 1976). With the
exception of Engelmann spruce, the density vs. distance
relationships were statistically significant atP < 0.01 or
greater (Tomback and others 1990).

For the Saddle Moimtain Burn, mean whitebark pine
regeneration density was not significantly different than
that for the Sleeping Child Bum, although somewhat
lower (table 1). The density was actually comparable to
that obtained for the ridgetop quadrats (west facing) in
the Sleeping Child Bum. In contrast, the densities for
subalpine fir and Engelmann spruce were much greater
than those in the Sleeping Child Burn. Lodgepole pine
was particularly abimdant as well. Scatterplots of quad-
rat density vs. distance from the upper seed source were
comparable to those from Sleeping Child (fig. 2).

0.14

0.10

r ••

0)
Q

0.05

0.30
0.20

0.10

P. albicaulis

• • •

A. lasiocarpa

•• •

• • • •

• ■ • • • •

• , • •••••• •

1000

2000

3000

Distance (m)

Figure 2 — Scatterplots of regeneration density
per quadrat versus distance from the whitebark
pine seed source for whitebark pine (top) and
subalpine fir (bottom) in the Saddle Mountain
Burn. From Tomback and others (1990).

Table 2 — Maximum age differences within clusters of whitebark

pine seedlings. Cluster sample size for the Sleeping Child
Burn = 87; for the Saddle Mountain Burn = 7. Based on
table 4, Tomback and others (1993)

Maximum age
differences

Sleeping Child

Saddle Mountain

Years

Percent

34.5

14.3

27.6

57.1

13.8

14.3

11.5

6.9

14.3

1.2

2.3

1.2

Whitebark pine regeneration showed a negative exponen-
tial distribution, and subalpine fir a similar but more lin-
ear negative relationship between quadrat density and
distance from the upper, unbiu-ned forest (density vs.
distance relationships significant at P < 0.001 for both).
Similar to findings for the Sleeping Child Bum, the En-
gelmann spruce scatterplot was nearly flat. For both
Douglas-fir and lodgepole pine, density increased with
distance from the upper, imbumed forest; major seed
sources for both species were at the lower edges of the
bum and, in the case of lodgepole pine, seeds from within
the bum. However, the relationship between density and
distance was not statistically significant for these latter
three conifers.

The distinctive negative exponential shape of the white-
bark pine regeneration curve probably results from two
factors: (1) whitebark pine seeds are only available at the
upper elevations of these burns, and (2) nutcrackers tend
to bury higher densities of seeds near parent trees. Ap-
parently, nutcrackers do cache seeds at long distances
from parent whitebark pine trees, because we found low
densities of regeneration up to 8 km from the whitebark
pine seed source (Tomback and others 1990). It could be
argued that the concomitant decrease in elevation in both
study areas was a confounding variable in the whitebark
pine density vs. distance relationship. However, recently,
I tested this possibility by examining whitebark pine re-
generation along a 3.4-km transect following an eleva-
tional isocline in the Sundance Bum of northern Idaho.
Like the Sleeping Child and Saddle Mountain Bxirns, the
whitebark pine seed source was restricted to one side of
the bum, and the resulting regeneration curve followed
a negative exponential distribution (Tomback 1992).

It appears that seed dispersal by nutcrackers has
provided whitebark pine an advantage over its shade-
tolerant competitor, subalpine fir, in the upper subalpine
of the Sleeping Child Burn. Because wind-dispersed fir
seeds from the upper seed source must travel against the
prevailing winds and cover a long distance, the regenera-
tion density of fir is lower than that for whitebark pine.
In contrast, in the smaller Saddle Mountain Burn, white-
bark pine and subalpine fir have comparable regeneration

195

densities. In this situation, dispersal of subalpine fir seeds
(and Engelmann spruce seeds) is favored both by the pre-
vailing winds from the west and the much smaller size of
the burn. Thus, for restocking burns, seed dispersal by
nutcrackers may provide whitebark pine an advantage
over wind-dispersed conifers when (1) the size of the burn
is large compared to distances traveled by wind-blown
seeds, (2) the seed source is unfavorably positioned with
respect to prevailing winds, and (3) the seed source is 1 km
or more from the perimeter of the burn.

Another consequence of seed dispersal by nutcrackers
is that seedling or sapling clusters occurred at more than
40 percent of the whitebark pine regeneration sites in both
the Sleeping Child and Saddle Mountain Burns (Tomback
and others 1990, 1993). Cluster sizes in the Sleeping
Child Burn ranged from two to eight stems with an over-
all mean of 1.9 stems per regeneration site (n = 455 total
sites). In the Saddle Mountain Burn, clusters ranged
from two to 10 stems, with an overall mean of 2.0 stems
per regeneration site (n = 164 sites). We aged two or more
cluster members from a cluster subsample from each of
the study areas (Tomback and others 1993). For the
Sleeping Child Burn, there were no differences in age
among seedlings within a cluster for 34 percent of the
clusters, differences of 1 year for 28 percent of the clus-
ters, and differences of 2 to 8 years within clusters for the
rest of the subsample (total n = 87 clusters; all seedlings
aged in 63 clusters) (table 2). Of the small number of
clusters sampled in the Saddle Mountain Burn, five clus-
ters (71 percent) had age differences of 0 to 1 year, and
the remaining two clusters had seedlings differing by
2 and 4 years, respectively (table 2). These resiilts strongly
suggest that asynchronous germination of seeds may oc-
cur in some clusters. This could be explained by delayed
maturation of some seeds or different sensitivities to envi-
ronmental conditions. Laboratory (Leadem 1986; Pitel and
Wang 1990) and experimental field studies (McCaughey
1993) indicate that embryo underdevelopment and de-
layed germination are tjrpical in whitebark pine seeds.
Delayed germination of seed caches would result in seed-
ling establishment following little or no cone production
and, thus, continuity of regeneration over time.

We compared age structiire (year of tree establishment
vs. time since fire) of regeneration for whitebark pine
and subalpine fir in both the Sleeping Child and Saddle
Mountain Burns to determine any important differences
(Tomback and others 1993). Both species appeared to be
similarly affected by environmental and ecological factors:
a disproportionately high percentage of regeneration was
established between 1977 and 1985, about 17 to 25 years
after fire (see fig. 1 in Tomback and others 1993). Thus,
major tree recruitment was episodic. A small increase
in mean January through August precipitation between
1977 and 1985 possibly contributed to this trend. If we
can generalize from these results, it appears that white-
bark pine and subalpine fir become established synchro-
nously after fire, albeit at different densities depending
on relative seed dispersal abilities.

REGENERATION IN THE GREATER
YELLOWSTONE AREA

Studies in the Greater Yellowstone Area following the
major fires of 1988 have provided comparative informa-
tion on patterns of seed dispersal, germination, and seed-
ling mortality in serai whitebark pine communities
(Tomback 1991, 1993). Unlike the large-scale, stand-
replacing burns described earlier for western Montana,
the Yellowstone fires created patches of burned areas of
different severities, often surrounded by unburned forest.

In 1990, my research assistants and I established a to-
tal of 275 permanent plots, each 20 m^, in two different
study areas. In the Cooke City study area, Gallatin Na-
tional Forest, we placed plots in four sample sites, repre-
senting the following treatments: dry, severely biirned
(50 plots); moist, severely burned (50 plots); dry, imbumed
(25 plots); moist, unburned (25 plots). Plot aspects in this
area ranged from 80° to 180°, and elevations rsmged from
2,680 to 2,745 m. On Mount Washbiirn, Yellowstone
National Park, we selected three sample sites, represent-
ing the following treatments: dry, severely burned (50
plots); moist, severely burned (50 plots); moist, moder-
ately burned (25). Aspects ranged from 250° to 360°,
and elevations ranged from 2,560 to 2,745 m. All conifer
regeneration was recorded for these plots in the siunmers
of 1990, 1991, and 1992; whitebark pine regeneration was
mapped for each plot. I briefly summarize the results
here (Tomback 1991, 1993).

Whitebark Regeneration Trends

Despite an abundant whitebark pine cone crop in 1989,
no whitebark pine regeneration occurred on any severely
burned or moderately burned plots in 1990 in either study
area (table 3). Only one plot in the moist, unburned treat-
ment had one newly germinated seedling cluster. Al-
though there was virtually no cone production in 1990,
all severely burned and moderately burned treatments
had some new whitebark pine regeneration in 1991, which
was probably the result of delayed germination of seeds
cached in 1989. Of the unburned treatments, again only
the moist plots experienced any new regeneration (table 3).

In 1992 there was some new whitebark pine regenera-
tion on the Cooke City moist, severely burned plots and
on the Mount Washburn moist, moderately burned plots.
By 1992, the latter sample site supported the highest den-
sity of whitebark pine regeneration of all sites (table 3).
Seedling survivorship from 1991 to 1992 was fairly high.
Those that survived to the summer of 1992 tended to be
shaded by other plants or forest debris for at least part
of the day. One cluster of seedlings that germinated in
1991 just outside a plot on the Mount Washburn moist,
severely burned sample site contained two new germinants
in 1992. This observation confirms the possibility that
seeds within caches may germinate in different years, as
illustrated by the ages of cluster members from western
Montana.

196

Table 3 — Whitebark pine regeneration densities for different plot

treatments in the Cooke City and Mount Washburn study
areas following the 1 988 Yellowstone fires

Mean density of regeneration sites^ per
Treatment 1990 1991 1992

Cooke City

Dry, severely burned

0.012

0.015

Moist, severely burned

.014

.010

Dry, unburned

.002

Moist, unburned

0.002

.010

.008

Mount Washburn

Dry, severely burned

.010

.011

Moist, severely burned

.021

.016

Moist, moderately burned

.030

.038

'A single regeneration site may support either a single seedling or a cluster
of seedlings arising from one seed cache.

Regeneration Trends of Other
Conifers

Trends in the regeneration of other conifers differed be-
tween the Cooke City and Mount Washburn study areas,
probably as a result of seed availability and seedbed con-
ditions. For example, no conifer regeneration, old or new,
occiirred on the Cooke City dry, severely burned plots in
1990; small niombers of Engelmann spruce and subalpine
fir occurred on these plots in subsequent years. On the
moist, severely burned plots, new spruce germinants
dominated in 1990 and both spruce and fir, new and old,
in subsequent years. Primarily new fir regeneration oc-
curred each year on the Cooke City dry, unburned plots,
with few older firs surviving; and, mostly spruce occurred
on the moist, unburned plots. Thus, spruce and fir regen-
eration predominated in the Cooke City area. Few lodge-
pole pine were apparent in the prefire forest or in the un-
bumed forest around the treatment sites.

In the Moimt Washburn study area, old and new lodge-
pole pine and new Engelmann spruce occiured in 1990 on
the dry, severely biirned plots. Older lodgepole pine and
some new spruce occurred in subsequent years. On the
moist, severely burned plots, new lodgepole pine and
spruce regeneration predominated in 1990, and older
lodgepole pine and spruce in 1991. In 1992 there were
munerous new germinants of both species and some sub-
alpine fir. High numbers of new spruce seedlings domi-
nated the moist, moderately burned plots in 1990, and
older and new spruce in subsequent years. Thus, lodge-
pole pine and spruce regeneration predominated in the
burned area sampled on Moimt Washbiirn. Lodgepole
pine had been an important species in the prefire forest
and was present, along with spruce, in imburned forest
not far fi-om the plot treatment sites. The high spruce re-
generation density on the moderately burned plots may
have been the consequence of a more favorable seedbed.

Although the numbers of whitebark pine germinants
were lower in most cases than those of the other conifer
species in both study areas, their siu-vivorship from 1991

to 1992 was relatively high. The microsites selected for
seed caching by nutcrackers and the hardiness of the
whitebark pine seedlings may result in lower mortality
compared to other conifers. We must follow the regenera-
tion patterns of whitebark pine and its forest associates
for several more years to test this idea.

CONCLUSIONS

The patterns of whitebark pine regeneration that we
observe are the consequence of the seed-dispersal behav-
iors and site preferences of nutcrackers, as well as their
tendency to bury multiseed caches. Nutcrackers bury
seeds in burned terrain soon after fire (for example, see
Tomback and Knowles 1989). Because whitebark pine
seeds vary in maturation rate, nutcrackers cache both de-
veloped and imderdeveloped seeds (Leadem 1986), result-
ing in germination over several different years from the
same seed crop.

Early postfire regeneration of whitebark pine may be the
consequence of rapid and widespread seed dispersal by
nutcrackers, suitability of the cache sites selected by nut-
crackers, and the hardiness of the seedlings themselves.

ACKNOWLEDGMENTS

The Intermountain Research Station, USDA Forest Ser-
vice, Ogden, UT, has funded all whitebark pine postfire
regeneration studies in western Montana, northern Idaho,
and the Greater Yellowstone Area. I thank Stephen Arno,
Jim Brown, and Bob Keane of the Intermountain Re-
search Station Fire Sciences Laboratory, Missoula, MT,
for logistic support and many helpful discussions concern-
ing this research. Steve Arno and Bob Keane reviewed
the draft manuscript. Lyn Hoffinann and Sharren Reuss
(Sund) provided assistance during the western Montana
studies, Kathy Carsey and Mary Powell worked with me
during the three field seasons of the Yellowstone study,
and Mary Powell, Jane Kees, and Mark Flower were field
assistants for the Idaho work. Support for attending the
SSPE Workshop was provided by the College of Liberal
Arts and Sciences, University of Colorado at Denver.

REFERENCES

Arno, S. F. 1980. Forest fire history in the northern
Rockies. Journal of Forestry. 78(8): 460-465.

Arno, S. F. 1986. Whitebark pine cone crops — a diminish-
ing source of vdldlife food? Western Journal of Applied
Forestry. 1(3): 92-94.

Arno, S. F.; Petersen, T. D. 1983. Variation in estimates
of fire intervals: a closer look at fire history on the Bit-
terroot National Forest. Res. Pap. INT-301. Ogden, UT:
U.S. Department of Agricult;ire, Forest Service, Inter-
mountain Forest and Range Experiment Station. 8 p.

Arno, S. F.; Hofif, R. J. 1989. Silvics of whitebark pine
{Pinus albicaulis). Gen. Tech. Rep. INT-253. Ogden,
UT: U.S. Department of Agriculture, Forest Service,
Intermountain Research Station, lip.

197

Finklin, A. I. 1983. Weather and climate of the Selway-
Bitterroot Wilderness. Moscow, ID: University Press
of Idaho. 144 p.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Leadem, C. L. 1986. Seed dormancy in three Pinus species
of the Inland Mountain West. In: Shearer, R. C, comp.
Proceedings — conifer tree seed in the Inland Mountain
West symposium; 1985 August 5-6; Missoula, MT. Gen.
Tech. Rep. INT-203. Ogden, UT: U.S. Department of
Agrictdture, Forest Service, Intermountain Research
Station: 117-124.

Lotan, J. E. 1976. Cone serotiny — fire relationships in
lodgepole pine. In: Lyon, L. J., comp. Tall Timbers Fire
Ecology Conference No. 14 and Intermountain Fire Re-
search Council fire and land management symposium:
proceedings; Missoula, MT: 267-278.

McCaughey, W. W.; Schmidt, W. C; Shearer, R. C. 1986.
Seed dispersal characteristics of conifers in the Inland
Mountain West. In: Shearer, R. C, comp. Proceedings —
conifer tree seed in the Inland Mountain West sympo-
sium; 1985 August 5-6; Missoula, MT. Gen. Tech. Rep.
INT-203. Ogden, UT: U.S. Department of Agriculture,
Forest Service, Intermountain Research Station: 50-62.

McCaughey, W. W. 1993. Delayed germination and seed-
ling emergence of Pinus albicaulis in a high elevation
clearcut in Montana, USA. In: Edwards, D. G., comp.
Seed dormancy and barriers to germination: proceed-
ings lUFRO symposium; 1991 April 23-26; Victoria, BC.
Victoria, BC: Forestry Canada. [In press].

Pitel, J. A.; Wang, B. S. P. 1990. Physical and chemical
treatments to improve germination of whitebark pine
seeds. In: Schmidt, W. C; McDonald, K. J., comps.
Proceedings — symposium on whitebark pine ecosys-
tems: ecology and management of a high-mountain
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
130-133.

Romme, W. H. 1980. Fire frequency in subalpine forests
of Yellowstone National Park. In: Fire history work-
shop: proceedings; 1984 October 20-24; Tucson, AZ.

Gen. Tech. Rep. RM-81. Fort Collins, CO: U.S. Depart-
ment of Agriculture, Forest Service, Rocky Mountain
Forest and Range Experiment Station: 27-30.

Tomback, D. F. 1978. Foraging strategies of Clark's nut-
cracker. The Living Bird. 16: 123-161.

Tomback, D. F. 1982. Dispersal of whitebark pine seeds
by Clark's nutcracker: a mutualism hypothesis. Journal
of Animal Ecology. 51: 451-467.

Tomback, D. F. 1991. Progress report on 1991 fieldwork
to U.S. Department of Agriculture, Forest Service, In-
termountain Research Station, Ogden, UT: patterns of
postfire regeneration of whitebark pine in the Greater
Yellowstone Area.

Tomback, D. F. 1992. Patterns of postfire whitebark pine
regeneration in the Sundance Burn of northern Idaho.
Unpublished data on file at: Department of Biology,
University of Colorado at Denver, Denver, CO.

Tomback, D. F. 1993. Progress report on 1992 fieldwork
to U.S. Department of Agriculture, Forest Service, In-
termountain Research Station, Ogden, UT: patterns of
postfire regeneration of whitebark pine in the Greater
Yellowstone Area.

Tomback, D. F.; Knowles, J. W. 1989. Observations on
nutcracker seed harvesting and caching on Mount
Washburn following the 1988 Yellowstone fires. Unpub-
lished data on file at: Department of Biology, University
of Colorado at Denver, Denver, CO.

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird-
dispersed pines. Evolutionary Ecology. 4: 185-219.

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Co-
evolution of nutcrackers and pines: implications for
forest regeneration. In: Schmidt, W. C; McDonald,
K. J., comps. Proceedings — symposium on whitebark
pine ecosystems: ecology and management of a high-
mountain resource; 1989 March 29-31; Bozeman, MT.
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department
of Agriculture, Forest Service, Intermountain Research
Station: 118-129.

Tomback, D. F.; Sund, S. K.; Hoffmann, L. A. 1993. Post-
fire regeneration of Pinus albicaulis: height-age rela-
tionships, age structure, and microsite characteristics.
Canadian Journal of Forest Research. 23: 113-119.

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of
the Clark's nutcracker and the pinon pine for efficient
seed harvest and dispersal. Ecological Monographs. 47:
89-111.

198

NEW TREND IN DENDROCHRONOLOGY:
1. THEORETICAL PRINCIPLES OF
REPROCHRONOLOGY

Vladislav N. Vorobjev

Abstract — Reviews previous research pertinent to tree growth
and cone-bearing relationships in Siberian stone pine (Pinus
sibirica) and other species. The ultimate objective is to secure
stable and moderately high cone production without significantly
jeopardizing tree growth. To reach this objective the author
points out potential avenues of research in tree growth dynamics
and reproductive activities, utilizing procedures used in
dendrochronology.

Dendrochronology was established as a science of the
growth dynamics of trees, the structure of the annual
ring, and the effect of internal and, especially, external
factors. The latter determined the development of den-
drochronology and a number of other special scientific
areas.

Generative processes have been studied separately and
dealt with by the corresponding branch of seed farming in
terms of the concept of the crop dynamics or crop cyclicity.
In doing so, novel independent results have been obtained
and, more important, premises have been created for com-
bining these findings with dendrochronology.

The correlation between the growth and generative de-
velopment of higher plants comprises a number of aspects
and has been studied for a long time. The study of herba-
ceous plants has lagged behind that of tree investigations.
Suffice it to say that not only have monographs on this
subject been lacking, but original publications contribut-
ing to this field have also been scarce. The only exception,
to the best of oiu- knowledge, has been the work done
in Tomsk by the laboratory of cone-bearing trees
(Goroshkevich 1989; Nekrasova 1974; Vorobjev 1983;
Vorobjev and Vorobjeva 1982; Vorobjev and others 1989).
However, the problem of the relation between growth and
cone bearing (in coniferous trees, in particular) is of great
importance in view of the necessity to increase cone crops.
While this problem is being solved in herbaceous plants
and fi-uit-bearing trees by an intense short-term shift of
metabolism toward activization of the generative pro-
cesses, this technique fails in coniferous trees because
of the long growth period inherent in them.

It is assumed on the basis of the foregoing consider-
ations that control of the sexual reproduction of trees

Poster presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Vladislav N. Vorobjev is Head of the Academy of Sciences, Siberian
Branch, Institute for Ecology of Natural Complexes, Academichesky pr. 2,
Academgorodok, Tomsk, 634055, Russia.

must be related to the determination of optimal correla-
tions between the generative and growth processes
through the long-term activization of the vital activity
of plants.

PREVIOUS RESULTS

The expedience of this approach was also confirmed
by the results of our previous investigations. The experi-
ments on the resin tapping of the Siberian cedar pine,
sometimes called Siberian stone pine {Pinus sibirica),
showed that the increase of cone crops by creating favor-
able conditions for cone bearing alone had no futiu'e be-
cause it would eventually resiilt in the weakening of
the growth processes and then of the generative ones
(Vorobjev 1974). Therefore, it appears necessary to look
for certain correlations between growth and cone bearing
that would make it possible to secure moderately high
cone crops on a long-term stable basis. It is this problem
that seems to be of great theoretical and practical interest
(Vorobjev and Vorobjeva 1980).

The study of the effect of the reproductive activity on
shoot growth and the shoot condition on subsequent cone
bearing provides insight into the relation between tree
growth and cone bearing. Observations have shown that
cone bearing depends on the previous shoot growth in
Scotch pine (Pinus sylvestris), English oak (Quercus sp.),
and apple (Malus sp.) trees (Minina 1954; Polozova 1957;
Pravdin 1950; Tsel'niker and Semikhatova 1957). It has
been pointed out that optimal shoot growth is required to
induce cone bearing. Of no less importance is the time it
takes to achieve growth, since the flower germs are initi-
ated when apical growth gets weaker or terminates. The
surface area of leaves also plays a significant role.

The effect of cone bearing on the current increment has
been the subject of a number of papers. In particular, the
diameter increment in coniferous trees was shown to de-
crease in the cone-maturing year (Chalupka and others
1975, 1976, 1977; Danilov 1953). However, the effect of the
decreased diameter increment on fiirther seminiference of
trees and metamere growth has not been examined yet.

The change in shoot growth due to cone bearing is the
net effect of sexualization and subsequent development
of generative organs at different stages of their maturity.
The dependence of the cambial growth of the shoot on its
sex has been described in literature (Nekrasova 1972;
Varnell 1976). However, no experimental evidence of the
role of sexualization and nutrition of the generative or-
gans, and of the effect of the nutrition of the 1- or 2-year
cones on the annual ring and its internal structure in the
growing shoot has been available (Vorobjev and Vorobjeva
1982).

199

RETROSPECTIVE METHOD

Good possibilities for the study of the effect of cone
crops on tree increment (both Hnear and cambial) have
opened up due to recent development of the method of ret-
rospective accoimt of the reproductive activity of trees
during their lifetime (see Vorobjev and others, these pro-
ceedings). Significantly, this procedure can also be used
to reproduce male "flowering" and the formation of a num-
ber of vegetative metameres, though for a shorter period
of time. In this case, shoots, branches, and crown become
major objects of inquiry in the investigation of the relation
between the growth and generative development of trees.

The research done into the growth-cone-bearing rela-
tionship in Siberian cedar pine in West Siberia and Gomy
Altai for the last 100 years or more (Vorobjev 1983) sug-
gests a new trend in the investigations made at the cross-
roads of the two sciences, biology of reproductive develop-
ment and dendrochronology.

This trend can be referred to as reprochronology, com-
prising both the cyclic fluctuations of the reproductive ac-
tivity and its relation to the cambial growth. Perhaps, the
word is not a happy choice from the terminology point of
view, but it generally implies sexual reproduction.

The time series of reproductive activity studied show
cycles of variable duration ranging from 3 to 28 years
or more. A 10- to 11-year reproductive cycle is less pro-
noimced than the growth cycle. The relation between the
crop dynamics and solar activity is feebly marked, and
seed cycles fall both at the peak and decay of the solar
activity. The reproductive activity is indirectly related to
the cambial growth and is delayed with respect to the lat-
ter by the time interval determined by the life cycle of one
generation and the growth aftereffect. The peak of the
siminiference activity generally falls at the descending
branch of the solar cycle and the period of the retarded
shoot growth. An increase in cone crops is related to the
previous solar activity and vigorous growth.

Thus it appears worthwhile to study further the struc-
ture of the time series of the reproductive activity and their
relation to the dynamics of growth of shoots, branches, and
trees, to evaluate the effect of the external factors using
new methods, and to develop prediction models. Also of
scientific importance are individual, interpopulational,
geographical, ecological, and other aspects of the relation
between the growth and the generative development of
trees as well as between the seminiference and the annual
ring of the shoot and tree. It should be noted that the
foregoing potential research areas suggest the use of the
techniques and procedures adopted in dendrochronology.

REFERENCES

Chalupka, W.; Giertych, M.; Krolikowski, Z. 1975. The
effect of cone crops on growth in Norway spruce {Picea
abies (L.) Karst.). Arbor. Kor. 20: 201-212.

Chalupka, W.; Giertych, M.; Krolikowski, Z. 1976. The
effect of cone crops in Scotch pine on tree diameter in-
crement. Arbor. Kor. 21: 361-366.

Chalupka, W.; Giertych, M.; Krolikowski, Z. 1977. Rela-
tion between specific gravity of wood in Norway spruce
(Picea abies (L.) Karst.): some growth parameters and
cone yield. Arbor. Kor. 22: 205-212.

Danilov, D. N. 1953. Effect of cone-bearing on tree-ring
structure in common spruce (Picea excelsa). Bot. J.
38(3): 367-377. [In Russian].

Goroshkevich, S. N. 1989. The dynamics of the micro-
strobile initiation in the male shoots of the Siberian ce-
dar pine as related to weather conditions. Ecol. 6: 33-39.
[In Russian].

Minina, E. G. 1954. Biological principles of flowering and
fruiting oak. Transactions of Inst, of Forest of USSR
Acad, of Sci. 17: 5-97. [In Russian].

Nekrasova, T. P. 1972. Biological principles of semi-
niference of Siberian cedar pine. Novosibirsk: Nauka.
272 p. [In Russian].

Nekrasova, T. P. 1974. The cyclicity of the cone-bearing of
Siberian cedar pine. In: Biology of seed reproduction of
coniferous trees of West Siberia. Novosibirsk: Nauka:
70-75. [In Russian].

Polozova, (L.) Y. 1957. Growth of oak shoots as related to
fruiting processes. Letters of Inst, of Forest of USSR
Acad, of Sci. 9: 9-17. [In Russian].

Pravdin, (L.) F. 1950. Sexual dimorphism in Scotch pine
(Pinus sylvestris L.). Transactions of Inst, of Forest of
USSR Acad, of Sci. 3: 190-201. [In Russian].

Tsel'niker, Y. L.; Semikhatova, O. A. 1957. On the rela-
tion between vegetative and generative stages of shoot
development in certain trees. Bot. J. 42(7): 1044-1054.
[In Russian].

Vamell, R. J. 1976. Correlation between branch charac-
teristics and production of female strobile in longleaf
pine. Forest Science. 22(2): 159-161.

Vorobjev, V. N. 1974. Growth and generative processes
in Siberian cedar pine under metabolic disturbance. In:
Biology of seed reproduction of coniferous trees of West
Siberia. Novosibirsk: Nauka: 140-170. [In Russian].

Vorobjev, V. N. 1983. Biological principles of compre-
hensive utilization of Siberian cedar pine forests.
Novosibirsk: Nauka. 254 p. [In Russian].

Vorobjev, V. N.; Vorobjeva, N. A. 1980. On the problem of
regulation of seed crops of coniferous trees. In: Selec-
tion, genetics and seed farming of trees as the basis for
forming of highly productive forests. Moscow: 391-395.
[In Russian].

Vorobjev, V. N.; Vorobjeva, N. A.; Goroshkevich, S. N.
1989. Growth and sex of Siberian cedar pine.
Novosibirsk: Nauka. 167 p. [In Russian).

Vorobjeva, N. A.; Vorobjev, V. N. 1982. The effect of fe-
male generative organs on the shoot growth in the
Siberian cedar pine. Lesovedeneye (For. Sci.) 1: 51-57.
[In Russian].

200

NEW TREND IN DENDROCHRONOLOGY:
2. METHOD OF RETROSPECTIVE
STUDY OF SEMINIFERENCE DYNAMICS
IN PINACEAE

V. N. Vorobjev

S. N. Goroshkevich

D. A. Savchuk

Abstract — Introduces a new method of retrospective analysis of
cone crop dynamics of Siberian stone pine {Pinus sihirica) for up
to 100 years before the present. For about the first 20 years, cone
abscissions are discernible on the exterior of the shoots and can
be used to reconstruct cone crop production, but after this they
are not identifiable. A method was developed utilizing cross-
sectional saw cuts to examine the interior of the branch shoots.
Traces of the abscissions of cones remain in the shoots and, con-
sequently, cone crops can be reconstructed for virtually the entire
life of the tree with the procedures described in this paper. This
method may also be useful for other conifers.

A retrospective study of the life activity of trees has al-
ways been of great interest to scientists. Retrospective
analysis has been widely practiced in dendrochronologic
and dendroclimatic studies focusing on the investigation
of tree rings whose structure and dimensions are impor-
tant indicators of the life activity of the tree organism.
The variations of tree increment observed are basically
related to the effect of exogenic factors, and cHmatic con-
ditions in particular. As to the endogenic factors and, pri-
marily, the reproductive activity, their contribution is rec-
ognized but not accounted for because of the difficulty of
estimating cone crops for long periods of time. In fact,
this aspect of activity of the tree organism is regarded
as the so-called "white noise" in a multifactor dendrochro-
nological analysis of the effect of different conditions.

The early information on the aimual dynamics of cone
crops has been obtained from direct counts of the number
of generative organs for several years. This technique is
fairly widely used even today. However, applications of
this technique are Hmited due to the necessity of long-
term observations. Naturally, scientists sought to develop
a method of retrospective study of the cone crop dynamics.
This method is based on the fact that the cones and fruits
either remain on the branches after the seed abscission
or fall out, leaving traces on the shoot bark. The seasonal

Poster presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

V. N. Vorobjev is Head of and S. N. Goroshkevich and D. A. Savchuk are
Members of the Academy of Sciences, Siberian Branch, Institute for Ecol-
ogy of Natural Complexes, 2, Acadcemichesky pr., Academgorodok, Tomsk,
634055, Russia.

character of the shoot growth in trees in the temperate
zone makes it possible to age each shoot and, conse-
quently, its generative organs.

The first information we are aware of on the tise of the
retrospective method of studying the crop dynamics can
be foimd in the work of A. Renvall (1912) for the Scotch
pine and N. Nesterov (1914) for some leaf -bearing trees
(oak, maple, hawthorn). The method was further devel-
oped and elaborated in great detail by Z. Trofimova
(1953), T. Nekrasova (1957), and P. Gorchakovsky (1958)
for different coniferous trees. A most comprehensive de-
scription of this method was given by A. Korchagin (1960).

The advent of the retrospective method has facilitated
significant progress in the investigation of crop dynamics.
The latter can be reproduced for the past 10-15 years on
the basis of the cone traces on the shoot bark. Special em-
phasis has been placed on Siberian cedar pine (sometimes
called Siberian stone pine) because of the great natural
and economic importance of its seed crops (Iroshnikov
1963; Vorobjev 1964, 1974). As a result, a cycUc charac-
ter of the fluctuations in the crops for different years was
revealed and the relations between the seminiference,
weather conditions, and tree grovTth were examined.
However, further advance of the investigations required
that the methodology be improved.

In our opinion, there are two very essential ways of im-
proving the method. First, work out a special technique
for reconstructing the stages of the development of each
generation from the initiation to the complete mattiration
of cones and to determine the number of cones that pre-
maturely abscissed in different stages of their develop-
ment. These data will make it possible to study the crop
dynamics and, in particular, its effect on the tree and
shoot growth and the annual ring structure. Second, ex-
tend the period of the retrospective account of crops from
10-15 years to 100 years or more.

The object of the present paper is to introduce a new
method of retrospective analysis of the seminiference dy-
namics of Siberian cedar pine, making use of the forego-
ing information.

CONE DEVELOPMENT

Cones in the Siberian cedar pine are initiated in late
July or early August (Nekrasova 1972). The germs of the
female cones hibernate in the form of undifferentiated

201

Figure 1 — Female cone traces on the shoot bark, x 6:
(a) cone that abscissed before pollination, (b) cone that
abscissed after pollination, (c) cone that abscissed during
maturation, (d) mature cone.

primordia surrounded by bud scales. The primordia are
located in the distal part of the winter terminal bud. The
development of the female cones begins in May. An active
linear growth of shoots occurs in the first half of June,
and the cones resemble large lateral buds in the distal
part of the growing shoot. Special observations show that
the development of some of the female cones ceases at this
period of time and they abort. For example, normally de-
veloping cones were bright green and 10-15 mm long in a
Siberian cedar pine forest near a settlement in the south
of the Tomsk region on J\me 6, 1990. Some of the female
buds began to burst, while others were of much smaller
size. In the latter case, a longitudinal section of a germ
showed that it had died in the differentiation phase. It
looked like a small (1.5-2.0 mm) brown protuberance.
After the degeneration of the germinal cone the bud dries
up and abscisses. Abortive female buds leave small (2.5-
3.0 mm) dark triangular traces (fig. la).

Pollination in Siberian cedar pine generally occurs in
the second half of July (Nekrasova 1972). After pollina-
tion the cones grow for some time until they are 2 mm
long. During this period (June-July) the cones may also
cease their development and abort. Traces of cones that
abscissed shortly after the pollination are larger (4-5 mm),
and more roundish than those of the abortive female buds
and are relatively smooth surfaced (fig. lb).

In the year of maturation both the cones and shoots
grow larger. The cones reach their ultimate dimensions
by the time of fertilization (Jvme 15-20) (Nekrasova 1972).
Therefore, the cones that absciss in the process of matura-
tion (June-August) leave the traces of the same shape and
size as normal mature cones do. The traces are large (up
to 7-8 mm) and oval with a funnel-shaped inner surface
(fig. Ic, d). Mature cones absciss in late August or early
September. The inner surface of their traces on the shoot
bark is covered with resin (fig. Id).

TRACE IDENTIFICATION

The identification of cone traces would require knowl-
edge of the morphostructure of the mature shoot and the
origin of its different parts. The annual shoot in large
cone-bearing branches in Siberian cedar pine consists of
two elementary shoots (Vorobjev and others 1989). The
first (spring) shoot develops from the winter bud in June.
The second (summer) shoot is initiated in June and is
found to enlarge at the beginning of Jvdy. Each elemen-
tary shoot has short shoots (brachyblasts) and lateral ex-
tending shoots. The length of the axis and the nvunber of
its metameres in the spring shoot are approximately ten
times as large as in the summer shoot. The female cones
are located in the distal part of the spring shoot. Small
lateral vegetative shoots usually grow at the same level

202

as the female cones. However, there may be no lateral
shoots. The main whorl of large lateral shoots is located
on the simimer shoot. It shoiild be noted that simimer
shoots in mature trees are formed when the spring shoot
bears cones. Therefore, the summer shoot can be regarded
as an indirect evidence of the presence of cones in the
spring shoot. The traces of mature cones on mature need-
less shoots will occur in the distal part of each annual
shoot under the main whorl of the lateral shoot.

The traces should be counted from the top of the branch
to yield the number of different types of traces on each
annual shoot. This procediire makes it possible to recon-
struct the entire dynamic cycle of the initiation, develop-
ment, and matiiration of cones (over 10 to 12 years). The
traces of abortive female buds on older shoots are de-
stroyed because of the secondary bulge of the shoot axis
and core cracking, which precludes their identification
and count. Large traces of mature cones persist for a
longer period of time, and they can be used for the recon-
struction of crop dynamics over approximately 20 years.

SAMPLING

Our experience shows that it is necessary to sample
10 branches from one tree. The branches selected must
be from different parts of the crown (top, middle, lower).
Each branch must be no less than 20 years old. To recon-
struct the djTiamics of the initiation and development of
cones on individual branches, one has to use branches of
mediimi size and ignore the traces of cones on the lateral
shoots. On the other hand, the seminiference dynamics
for the tree or stand as a whole can be followed by using
cone-bearing branches of different size and, most impor-
tant, by taking into account the traces on small lateral
branches.

As the number of annual rings increases, the traces
of cones on the bark in the cone-bearing branches of trees
become indiscernible. Therefore, a procedure was devel-
oped for the identification of such traces on the cross-
sectional saw cuts of branches (Vorobjev 1979). The saw
cuts of the bases of whorls show that the traces of cones
remain within shoots and are quite discernible, which
makes it possible to reconstruct the crop dynamics virtu-
ally for the entire Ufe of the tree. Thus, the time scale of
the proposed procedure can be extended.

The saw cuts must be made 1 cm beneath the point
where the longitudinal axis of the branch intersects the
axes of the lateral shoots. Then they are paper-sanded
and examined with a stereomicroscope x 16. The saw cuts
must begin with the whorls where the external traces on
the bark are fairly visible. The identification of the inner
traces should also begin with those whorls. This is espe-
cially important for the differentiation of different periods
of cone development.

TRACE CHARACTERISTICS

The cone traces on the saw cuts of branches can be
described as follows: The trace of the cone that abscissed
before pollination (the trace of the abortive female bud)
has the form of a brown ray with a uniform width of

0.3±0.03 mm and a length of 1.6±0.2 mm. It runs from
the pith of the shoot to its bark at the backgrotmd of a
yellow saw cut wood (fig. 2a). The trace of the cone that
abscissed after pollination and that of the immature cone
are half as large edgewise as the trace of the cone that
abscissed before pollination. The latter trace (fig. 2b) in
contrast to the former (fig. 2c) has a bulge at the end (1.7
times as large as the ray width) or is slightly forked. The
trace of the mature cone (fig. 2d) is twice as large as that
of the immature cone. Note that there may be no ray.
The trace of the mature cone differs greatly from all the
other traces (see fig. 2d and figs. 2a, b, c) by an oval or
triangular bulge (2.4+0.3 mm wide) at the ray end. Such
traces are often accompanied by a specific deformation of
the tree rings as a result of rupture of phloem and xylem
tissues of the shoot by a great n;unber of cones. Thus the
more mature the cone the brighter and longer is its trace
on the saw cut. The bulge is also larger.

The traces of cones differ from the traces of the lateral
shoots and brachyblasts on the cross-sectional saw cuts.
The trace of the lateral shoot (fig. 2e) is wide and gener-
ally runs toward the bark to join the shoot axis. The trace
of the cone does not reach the bark. The trace of brachy-
blast (fig. 2f) is very similar to the trace of the cone that
abscissed before pollination (the abortive female bud), but
the former trace is much longer (2.5+0.2 mm) and its end
grows narrower. The main thing is that the trace of
brachyblast is surrounded by a brighter layer of wood
(3.8±0.3 X 0.7±0.1 mm), which is visible to a naked eye.

USE OF METHODS

We feel that use can be made of methods based on the
difference in the density of the cone traces and that of the
surrounding wood (roentgenography, densitography, laser
techniques, etc.) in the cases where the traces on the saw
cuts are dim (especially the traces of the prematvirely
abscissed cones), or else, if the saw cut has been made in
a wrong way. In particular, soft x-ray roentgenography
of a 5- to 7-mm-thick saw cut of a branch (the time of expo-
sure was 6 minutes) made it possible to determine unam-
biguously the structure of tree rings, cone traces, and those
of other metameres (Vorobjev 1979). The above techniques
provide docvmientary evidence for subsequent analysis.

This ability to distinguish the internal traces must be
practiced using the branches with fairly visible external
cone traces on the bark. The selection of the branches
for the analysis must be made in the female zone of the
crown. The branches with one top are most suitable for
examining the relationship between the seminiference
and the shoot growth. For the study of crop dynamics,
on the other hand, the branches with a few tops must be
used. These branches often occur in the lower part of the
female zone.

Sample branches of the tree must be taken from three
age groups (30, 60, and 100 years old). The extreme val-
ues of the crop capacity (the number of cones on one
shoot) are omitted in the case of young branches for the
first few years before cone bearing and in the case of old
branches for the last tens of years after cone bearing. The
extremes are ignored because of the absence of cone

203

Figure 2 — Traces of generative and vegetative
organs on cross-sectional saw cuts of branches.
X 3: (a) cone that abscissed before pollination,
(b) cone that abscissed after pollination, (c) cone
that abscissed during maturation, (d) mature
cone, (e) lateral shoot, (f) brachyblast.

traces. The information is taken into account within the
number series, for it reflects nonseed years. The traces
are hardly visible at this time because of the predominant
abscission of 1-year-old conelets caused by early frosts or
the lack of pollination.

No less than three to five 200- to 250-year-old trees of
the first or second Kraft's class are to be selected from one
sample area for the study of sample branches.

A retrospective study of the seminiference of Siberian
cedar pine on the basis of the proposed method makes it
possible to assess the effect of the reproductive activity of
trees on the annual ring structure and size, to follow the
crop cyclicity, to elaborate the ways of the long-term and
super-long-term forecasting of crops, and to construct an
ontogenetic model of the relationship between the growth
and seminiference of shoots. This method appears appli-
cable to other coniferous plants and, first and foremost,
to Pinaceae.

REFERENCES

Gorchakovsky, P. L. 1958. New issues in the method of in-
vestigation of the seminiference dynamics of coniferous
trees. Bot. Zhourn. 43(10): 1445-1459. [In Russian].

Iroshnikov, A. I. 1963. The cone-bearing of Siberian cedar
pine in western Sayani. In: Cone-bearing of Siberian
cedar pine in eastern Siberia; Moscow: Publ. of USSR
Academy of Science: 104-119. [In Russian].

Korchagin, A. A. 1960. Methods of account of seminiference
of trees and forest communities. In: Field Geobotany;

Moscow-Leningrad: Publication of USSR Academy
of Science: Vol. 2: 41-133. [In Russian].

Nekrasova, T. P. 1957. On the methods of investigation
of cone-bearing dynamics in coniferous trees. Izv. Vost.
Fil. Akad. Nauk SSSR. 6: 138-145. [In Russian].

Nekrasova, T. P. 1972. Biological principles of seminif-
erence of Siberian cedar pine. Novosibirsk, Nauka:
272 p. [In Russian].

Nesterov, N. S. 1914. On the problem of the methods of
investigation of cone-bearing of trees. Forest-Industrial
Vestnik. Vol. 26. [In Russian].

Renvall, A. 1912. Die periodischen Erscheinungen der
Reproduction der Kiefer an der polaren Waldgrenze.
Acta Forest Fenn. 22(3): 130 p.

Trofimova, Z. I. 1953. Determination of pine crops by bio-
logical method. For. Econ. 1: 67-69. [In Russi£Ui].

Vorobjev, V. N. 1964. The cone-bearing of Siberian cedar
pine in altitude subzones of north-eastern Altai. Izv.
Sib. Otd. Akad. Nauk SSSR, Ser. Biol.-Med. Nauk.
12(3): 86-90. [In Russian].

Vorobjev, V. N. 1974. Peculiarities of cone-bearing of
Siberian cedar pine in mountain conditions. In: Biology
of seed reproduction of coniferous trees of western
Siberia. Novosibirsk, Nauka: 15-70. [In Russian].

Vorobjev, V. N. 1979. A method of retrospective study
of seminiference dynamics in Pinus sibirica Du Tour
(Pinaceae). Bot. Zhourn. 64(7): 971-974. [In Russian].

Vorobjev, V. N.; Vorobjeva, N. A.; Goroshkevich, S. N.
1989. Growth and sex of Siberian cedar pine.
Novosibirsk, Nauka. 167 p. [In Russian].

204

1 ^ \ \

Importance to Wildlife

International Workshop
St. Moritz 1 992

^ — ^

205

NUTRITIVE VALUE OF WHITEBARK
PINE SEEDS, AND THE QUESTION OF
THEIR VARIABLE DORMANCY

Ronald M. Lanner
Barrie K. Gilbert

Abstract— Seeds of whitebark pine (Pinus albicaulis) are eaten
year-round by their disperser, Clark's nutcracker (Nucifraga
Columbiana), and may be the only food that is fed to their young.
They are also preferentially larder-hoarded by the red squirrel
{Tamiasciurus hudsonicus); and are eaten in large quantities by
black bears (Ursus americanus) and grizzly bears {U. arctos), es-
pecially prior to hibernation.

Analyses show whitebark pine seeds have a mean seed weight
of 119 mg, of which kernels comprise 52.5 percent. Kernel content
was approximately 21 percent carbohydrate, 21 percent protein,
52 percent fat, 3 percent ash, and 3 percent water. Major mineral
nutrients present were Cu, Zn, Fe, Mn, Mg, and Ca. Sixteen amino
acids were present, with glutamic acid, lysine, and arginine espe-
cially abundant. Major fatty acids were linoleic, oleic, and y-linolenic,
with lesser amounts of 11 others. The unsaturated fraction was
92 percent.

The variable period of dormancy is analyzed in terms of its con-
tribution to fitness in a variable environment, its phylogenetic
distribution within Pinus subg. Strobus and subsect. Cembrae,
and its effect in retaining nutcracker populations by broadening
the food base. The food base is broadened by the interaction of
memory duration, seed crop frequency, and foraging on buried
dormant seeds whose location is betrayed by presence of germi-
nants. This may explain the excessive caching behavior of nut-
crackers and may reduce the frequency of irruptions during times
of food shortage. Even when seed crops are sporadic, variable dor-
mancy can make seeds available to nutcrackers in most years, and
as early as the first year after dispersal birds can forage on seeds
cached by other birds.

Whitebark pine (Pinus albicaulis Engelm.) is the only
North American member of Pinus subsect. Cembrae, the
stone pines. The species is found in subalpine habitats in
the Northern Rocky Mountains of Canada and the United
States, and in the Cascade Mountains-Sierra Nevada sys-
tem (Critchfield and Little 1966). Because whitebark pine
usually grows in inaccessible areas and has relatively low
commercial value, it was long neglected by researchers.
But in recent years, due largely to interest in corvid-pine
mutualisms stimulated by a seminal review (Turcek and

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Ronald M. Lanner is Professor, Department of Forest Resources, and
Barrie K. Gilbert is Senior Scientist, Department of Fisheries and Wildlife,
both at Utah State University, Logan, UT.

Kelso 1968), it has received greatly increased attention.
Much of this attention has focused on the important role
of whitebark pine seed in the biology of animals that in-
habit the whitebark pine forest.

It is the large Upid-rich "nut" of this pine that attracts its
seed disperser, Clark's nutcracker (Nucifraga columbiana)
(Tomback 1978). Evidence indicates that the pine can reU-
ably regenerate only from seeds cached in the soil by its nut-
cracker mutualist (Hutchins and Lanner 1982), and that
these seeds are the sole nutritional resource for nutcrackers
throughout their juvenile stage (Vander Wall and Hutchins
1983). Further, it has been shown that whitebark pine
seeds are a preferred food of the red squirrel (Tamiasciurus
hudsonicus), which larder-hoards great quantities of cones
on and below the groimd (Hutchins and Lanner 1982). Both
black bears (Ursus americanus) and grizzly bears (U. arctos)
raid the squirrel middens in order to gorge themselves with
pine nuts both before and after hibernation (Kendall 1983;
Mattson and Jonkel 1990). Thus the whitebark pine seed
is of obviously great nutritional importance to at least five
organisms of its ecosystem: Clark's nutcracker, the red
squirrel, black and grizzly bears, and whitebark pine itself,
for whose benefit the seed originally evolved. It is therefore
of more than passing interest to know the nutrient content
of this valued seed.

Pubhshed analyses of whitebark pine nuts have stressed
their high caloric value, which is mainly a function of high
lipid content. Emphasis on caloric content appefirs to be a
response to the current ecological fixation on energetics of
food resources, almost to the exclusion of other food charac-
teristics. Thus Lanner (1982) has documented the parti-
tioning of energy in nuts and other cone components, and
Hutchins and Lanner (1982) have reported changes in the
caloric value of whitebark pine seeds as the seeds mature,
but other aspects of the nutritional qualities of these seeds
have been neglected.

Another characteristic of the whitebark pine seed that
requires attention is its variable dormancy. Seeds of several
Cembrae species have been reported to delay germination
until the second or third year following dispersal (USDA
FS 1974). This has been attributed to the immaturity of
embryos at the normal germination date, and has been
blamed for low germination rates of whitebark pine seeds
in laboratory tests (Pitel and Wang 1990). Kozlowski and
Gunn (1972) have suggested that germination delays ensure
estabhshment "even though the early germinants failed to
survive severe environmental stresses such as droughts or
severe frosts," but they offer no data to support this intu-
itively attractive idea. According to McCaughey (1992),
whitebark pine seeds of a cohort displayed germination

206

dming the first (11 percent), second (45 percent), third
(11 percent), and fourth (percent not specified) postdis-
persal years.

The objectives of this paper are to document the nutri-
tional characteristics of whitebark pine seeds, and to sug-
gest some possible ramifications of the variable dormancy
habit.

SEED COMPOSITION ANALYSES

Analyses were made of whitebark pine seeds of the 1989
crop fi-om around Gardiner, MT (2,635 m elevation), provided
by Dr. Ward McCaughey of the Intermountain Research Sta-
tion (USDA Forest Service), Bozeman, MT.

Protein determination was made by the Kjeldahl method,
fat determination by the chloroform methanol method of
Hubbard (1977), and mineral assays by atomic absorption
by Drs. Arthiir Mahoney and Dejia Zhang, Department of
Nutrition and Food Sciences, Utah State University. Five
samples were used for these determinations. Amino acid
analyses of nut protein were made by Dr. Rod Brown, De-
partment of Nutrition and Food Sciences, Utah State Uni-
versity, using a Beckmann 6300 amino acid analyzer. Four
nms were made: data presented here are the means of those
runs. Fatty acid composition of kernel fats was provided by
Laurence G. Cool of Dr. E. Zavarin's laboratory at the For-
est Products Laboratory, University of Cahfomia, Richmond,
CA, and was determined by gas chromatography.

Results

Kernel Composition — Mean values of 20-seed samples
yielded the following seed coat versus kernel (female game-
toph5^e plus embryo) breakdown: seed coat, 47.5 percent,
0.0568 g; kernel, 52.5 percent, 0.0626 g. Kernels consisted
of water (3.5 percent), ash (2.8 percent), carbohydrates
(21 percent) [calculated value], protein (21 percent), and

Table 1 — Amino acids identified in kernel
protein of whitebark pine in order
of abundance (mean of four runs)

Amino

Nanomoles

acid

per mg

Glutamic acid

0.729

Lysine

.666

Arginine

.554

Aspartic acid

.422

Glycine

.386

Valine

.380

Alanine

.373

Leucine

.369

Serine

.363

Proline

.352

Isoleucine

.245

Phenylalanine

.192

Tyrosine

.168

Threonine

.147

Histidine

.126

Methionine

.038

Table 2 — Comparative abundance of amino acids identified in
kernel protein of three pine species^

Amino
acid'

P. albicaulis

P. edulis

P. pinea

^1 1 1

1 An

0 /

ft7

1 vo
LYb

y 1

1 Q

Ano

1 nn

1 UU

1 nn

ACD

RQ
OO

vaLY

til

VAL

Al A
MLA

1 CI 1

O /

CCD
OtPl

OO
r r

rnU

OQ.

loU

'>A

OCi

Due

rn t

1 Yn

H n

i»

THR

HIS

MET

CYS

TRY

'Species and references: P. albicaulis (this paper), P. edulis (McCarthy and
Matthews 1984), P. pinea (McCarthy and Matthews 1984).

^Amino acids not previously mentioned; CYS = cystine, TRY = tryptophan.

fat (52 percent). Atomic absorption analysis disclosed the
following mineral composition, in micrograms per gram of
ash: Cu, 17.5; Zn, 66.9; Fe, 35.7; Mn, 132; Mg, 17.5; Ca, 51.9.

Amino Acids — Sixteen amino acids were identified in the
whitebark pine kernel samples. The most abimdant were
glutamic acid, lysine, arginine, and aspartic acid (table 1).
These ranged from 0.42 to 0.73 nmol/mg. Also abxmdant,
ranging from 0.35 to 0.39 nmol/mg were glycine, valine, ala-
nine, leucine, serine, and proline. Lesser quantities were
found of isoleucine, phenylalanine, tyrosine, threonine, and
histidine; and least of methionine (table 1).

Amino acid quantities have been expressed in the Htera-
ture in various ways. Therefore, to compare the amino acids
of whitebark pine with those of other species, we have given
the most abimdant amino acid of each species the index
value of 100, and aU others are scaled appropriately within
each species. Thus, while the most abimdant amino acid
(100) of whitebark pine is glutamic acid, that oiPinus edulis
and P. pinea is arginine (table 2).

Fatty Acids — Fifteen fatty acids were identified on the
chromatograph. The most abimdant of these were the poly-
unsaturated linoleic acid and monounsaturated oleic acid,
which comprised 46.9 percent and 30.6 percent of the total
fatty acid methyl esters, respectively. The other components
are present in much smaller proportions (table 3). The total
fraction of imsaturated fatty acids was 92.0 percent, 32.1
percent monoimsaturated, and 59.9 percent polyunsaturated.

Discussion

Due to differences in analytical techniques, and variation
in the multipHer used in determining protein content, the
pubUshed fractions of fats, proteins, carbohydrates, and
ash in nut kernels are quite variable.

207

Table 3 — Fatty acids identified in kernel fat of
whitebark pine in order of abundance

Total fatty acid

Fatty acid

methyl esters

Percent

Linoleic

46.9

Oleic

30.6

y-Linolenic

9.9

Palmitic

4.2

Stearic

2.7

1 1 -Eicosenoic

1.4

y-Linoleic

1.2

5-, 11-, 14-Eicosatrienoic

1 1 -, 1 4-Eicosadienoic

Arachidic

Linolenic

DOl lc?l IIU

Vaccenic

5-, 1 1 -Eicosadienoic

Unknowns

Total

100.0

Whitebark pine's protein content of 21 percent compares
with pinyon pine values in the range of 10 percent (Pinus
monophylla) to 19 percent (P. cembroides) (Lanner 1981);
25 percent in P. sabiniana, and 31 percent in P. pinea
(Farris 1983).

The proteins in pine nuts are almost entirely seed storage
proteins of the megagametophyte. Following the hydrolysis
of these proteins, the component amino acids are exported
into the growing points of the embryo (Lammer and GifiFord
1989). GifiFord (1988) has shown there is considerable varia-
tion in electrophoretic characteristics of seed proteins of
several species (including P. albicaulis); and Schirone and
others (1991) have shown such variation to have taxonomic
significance. It is therefore surprising to see the almost-
identical gmaino acid profiles of P. pinea and P. edulis seed
proteins (table 2). These species are in separate subgenera
of Pinus (Critchfield and Little 1966) and would be expected
to have proteins of quite different composition.

Foiu" common amino acids were missing fi-om whitebark
pine seed protein. Hydroxyproline's absence can probably
be attributed to a lack of that amino acid in embryo and
storage protein, and its restriction to seed coats and peri-
carp material (Van Etten and others 1967). Tr3T)tophan is
generally a minor component of plant proteins (White and
others 1959) and can be destroyed in analysis (Farris 1983).
Neither cystine nor cysteine were identified in whitebark
pine, though the former is found in small quantities in
P. edulis and P. pinea (table 2). The lack of cystine and the
low quantity of methionine indicate that sulfur-containing
amino acids are of little importance to development of white-
bark pine seedlings.

Perhaps the most striking difference between seed protein
of whitebark pine and the other species in table 2 is its great
abundance of lysine. Lysine, an essential amino acid for
human growth, is notoriously low in cultivated members of
the Graminae. Feeding experiments conducted with dark-
eyed juncos iJunco hyemalis) by Parrish and Martin (1977)

have established the essentiality of lysine in maintaining
a positive nitrogen balance in that bird. Additional amino
acids found essential for poultry are arginine, glycine, £ind
histidine, all of which are found in whitebark pine nuts
(table 1). In the absence of data specific to the genus Nuci-
fraga, we speculate that these nuts contain all the amino
acids essential to nutcrackers.

According to Van Etten and others (1967) the nutrition-
ally essential amino acid requirements for swine are similar
to those of humans. If the requirements for bears are also
reasonably similar, then whitebark pine protein, as well as
its fat, may be of great dietary significance for bears. Bears
are more likely to satisfy their protein needs firom a wider
variety of foodstuffs than nutcrackers, however.

The fat content of whitebark pine nuts in this study (52
percent) is qmte similar to that of other stone pines as deter-
mined by Tikhomirov (1939): P. cembra, 59.9 percent and
P. pumila, 59.4 percent. Fat values for P. sibirica vary over
the range 51 to 75 percent (Shimanjoik 1963). The fat con-
tent of pinyon pines (Lsmner 1981) ranges fi'om 23 percent
in P. monophylla to 62 to 71 percent in P. edulis. Farris
(1983) reports that P. sabiniana kernels contain 49 percent
of fat, and P. pinea contains 47 percent.

Tikhomirov (1939) reported the fat content of P. koraien-
sis to be 52.5 percent, but recent work by Yoon and others
(1989) reports a crude lipid fi'action of 72.5 percent, which
reduced to 70.4 percent when purified.

The most notable features of the whitebark pine seed's
fatty acids are the preponderance of oleic (30.6 percent) and
linoleic (46.9 percent) acids. These are both unsaturated
fatty acids, and it appears they are commonly the major
ingredients of the seed fatty acids (f.a.) of Pinus. Thus they
comprise 28.4 percent and 47.9 percent of the f.a. of P. kor-
aiensis, 40.0 percent and 41.9 percent of the f.a. of P. edulis
(R. M. Lanner, impubl. data), and 42.5 percent and 42.0 per-
cent of the f.a. of P. monophylla, respectively (R. M. Lanner,
impubl. data). According to Yoon and others (1989) the f.a.
of P. koraiensis are 8.75 percent saturated, 29.41 percent
monounsaturated, and 61.64 percent poljomsaturated. Pinus
edulis and P. monophylla have over 80 percent polyunsatu-
rates. The nut of P. sabiniana was found by Farris (1983)
to be 50.5 percent oleic and 45.2 percent linoleic acids —
over 95 percent polyimsaturated.

In addition to containing very high ratios of imsaturated
to saturated fatty acids, pine nuts are also fi'ee of cholesterol
(McCarthy and Matthews 1984). Yoon and others (1989)
point out that P. koraiensis nuts have long been used in
Korean medicine, and the nut oil of P. cembra had medici-
nal uses in Poland during the 17th century (Bialobok 1975).
Yoon and others point out that P. koraiensis seeds contain
omega-5 fatty acids, the functions of which in the himian
body are unknown. They suggest that these f.a. may be
beneficial to health, and should be studied in relation to
lipid and glucose metabolism.

VARIABLE DORMANCY

The variable dormancy that characterizes whitebark
pine seeds has been viewed as a bothersome habit that
makes the nurseryman's task more difiScult. However, it
could well be a trait that has an important biological func-
tion and that has been naturally selected to perform that

208

Table 4^Number of years during a 15-year period in which seeds will be germinating, under assumptions of different seed-crop intervals, and
of dormancy varying from 0 to 3 years^

Period of seed

dormancy Interval between successive seed crops

'Seeds dormant for 0 years all germinate the spring following dispersal. Seeds domiant more than 0 years germinate in part in the spring following disperal; and
in part during each additional spring for 1 , 2, or 3 years.

function. Here we postulate and examine several possible
evolutionary functions for variable dormancy.

Increased Fitness

According to this hypothesis, the conditions met by ger-
minating seeds may vary greatly from year to year. There-
fore a cohort of seeds dispersed in fall of the year n may be
confronted by disadvantageous — or even catastrophic —
conditions in spring of n+1. Such conditions could result
from severe frost or drought (Kozlowski and Gunn 1972).
If so, it might be to the plant's advantage to delay germi-
nation of some seeds until spring of n+2 or even n+3 or be-
yond. If the temporal spacing of seed crops and the lengths
of the variable dormancy periods are matched properly, par-
tial seed cohorts could be germinating nearly every year.
Under such a sceneirio there would be a high probability of
some seeds germinating during the years of most favorable
conditions.

As an example, assimie a pine stand bearing seeds in the
first year of an arbitrarily chosen 15-year period. If seed
crops reoccur every 4 years (in years 5, 9, and 13 of this
15-year period), and if germination occiu-s only in the spring
foUovdng dispersal (spring of years 2, 6, 10, and 14), then
there will be 4 years during the 15-year period when seed-
lings can become established (table 4). However, if some
of those seeds remain dormant for variable lengths of time,
germinating not only in the spring following dispersal, but
in the next three springs as well, seedlings can become es-
tablished during all 15 years of the period (table 4). Even
if seed crops were spaced at 10-year intervals, this level of
variable dormancy would produce seedlings in 8 of the 15
years. Thus it is apparent that variable dormancy can neu-
tralize the effects of infrequent crop years by creating a
steady stream of germinating seeds during and between
crop years.

If variable dormancy increases whitebark pine's fitness by
increasing the frequency of germination years, one might ex-
pect associated conifers — especially other pines — ^to show the
same dormancy behavior. Apparently, however, this seems
not to be the case. If we assume that stratification treat-
ments prescribed to overcome seed dormancy are an index
of a seed's behavior in nature, it is apparent that the subal-
pine conifers of western North America show no consistency
(table 5). This does not mean that whitebark pine derives
no benefits from being able to provide germinants in most

years, but perhaps that only whitebark pine (and maybe
P. monticola) enjoys those benefits. Apparently variable
dormancy is not a requirement for success of whitebark
pine's associates in the subsdpine zone. In fact, variable
dormancy might be detrimental to these other conifers,
because, except for P. flexilis, they have wind-dispersed
seeds that lie on the soil surface, where they are vulner-
able to predation. Buried seeds of whitebark pine (and lim-
ber pine) would be more likely to survive an additional year
or more of "waiting" to germinate.

If variable dormancy is a genetically controlled charac-
ter, it would be important to learn its mode of inheritance.
For example, do all the seeds of a tree exhibit the same
dormancy behavior, as would be expected in a maternally
transmitted characteristic of seed-coat or megagametophytic
tissue; or do they segregate, as would be expected if inher-
itance from both parents is manifested in variable embryo
behavior? If the latter, is genetic control due to major gene
effects, or is control achieved multigenically? Seed tests
would need to be made with the products of controlled
crosses to unravel this and similar questions, and careful
anatomical studies would have to be made to better imder-
stand embryo growth and development.

Table 5 — Recommended^ cold stratification periods for seeds of
subalpine conifers of western North America and their
overlap with that of Pinus albicaulis

Overlap of

Stratification P. albicaulis

Species period stratification period

Days Percent

Pinus albicaulis

90-120

P. monticola

30-120

100

P. flexilis

21 - 90

P. aristata

P. balfouriana

P. contorta var. latifolia

30- 56

P. contorta var. murrayana

20- 30

Tsuga mertensiana

Picea engelmannii

Abies lasiocarpa

A. amabilis

21 - 28

A. magnifica

28- 42

In USDA FS1974.

209

Phylogenetic Legacy

Perhaps whitebark pine exhibits variable dormancy sim-
ply because the trait is conmion to a group of related species
that includes whitebark pine. To determine this, we have
displayed in table 6 the available stratification requirements
of pines of Pinus subg. Strobus by subsection.

Species characterized (in USDA FS 1974) as "suspected
of having immature embryos at the time of collection" or of
exhibiting "extreme dormancy" are identified by superscript
nvunbers. The densest concentration of superscripts is in
the stone pines — subsection Cembrae — in which all species
require more than 60 days of stratification to germinate
(table 6). Thus it seems likely that variable dormancy is
not just an albicaulis trait, nor a broad Strobus trait, but
a subsection Cembrae trait. This is consistent with my ear-
lier suggestion (Lanner 1980) that the Cembrae pines are
of monophyletic origin.

Nutcracker Retention

The importance of variable dormancy may lie in its ef-
fects on whitebark pine's seed disperser, Clark's nutcracker.
Whitebark pine depends on the nutcracker for seed dispersal
and seedling establishment (Hutchins and Lanner 1982),
so it is in the long-term interest of the pine to retain nut-
cracker populations in its ecosystem. We suggest that vari-
able dormancy can have that effect.

The following scenario illustrates our argument. A mast
year occurs in whitebark pine (year n). The local nutcracker
population frantically harvests and caches several times
as many seeds as are required to ensure its own survival
and that of its offspring imtil the next fall (Hutchins and
Lanner 1982; Tomback 1978; Vander Wall and Balda 1977).
Throu^out the fall, the nutcrackers recache their buried
seeds (Hutchins and Lanner 1982). The next summer, in
year n+1, some seeds germinate. The birds now have three
sources of food — their own caches, fi'om which they continue
to feed; the seeds now germinating and therefore visible;
and the still-dormant seeds fi'om the same caches, whose
locations are betrayed by the n+1 germinants (Vander Wall
and Hutchins 1983). Each bird now knows not only the
locations of all its own caches, but also the caches made
by other birds in which n+1 germinants were found. Even
if there is no crop in the second year, the bird's 9-month
memory (Balda 1992) of cache location will allow it to sur-
vive until the third year, when a new crop may be present.
Again, sprouting seeds of n+2 will betray the locations of
remaining caches. If there is again no crop, or only a light
one, the nutcrackers may be able to survive by recaching
the dormant seeds they have found associated with germi-
nating seeds. This assumes that another 9-month period
of cache remembrance follows each caching event. Due to
the overlapping of periods of availability of current-crop
seeds and seeds of the previous crop still lying dormant up
to 3 years in caches, plus those germinating, nutcrackers
should be able to remain in situ except in years following
several successive crop failures.

Such a scenario could not occur with seeds that all
sprouted in n+1. With variable dormancy it appears to be
a strong possibility; however, intensive observation and
additional data will be required to determine whether it

Table 6— Recommended^ cold stratification periods for seed of

some pines in subgenus Strobus, and their overlap with
that of Pinus albicaulis

Species

Stratification
period

Overlap of
P. albicaulis
stratification period

Days Percent

(Pinus)

albicaulis^ 90-120 —

subsection Strobi

flexilis^ 21-90 3

monticola^ 30-120 100

strobus 60 0

strobiformis^ 60-120 100

lambertiana^ 60- 90 3

peuce 0-60 0

parviflora^-^ 90 3

griff ithii 0-15 0

subsection Balfourianae

aristata 0 0

balfouriana^ 90 3

subsection Cembroides

cembroides 0 0

edulis 0 0

monophylla^ 28- 90 3

quadrifolia 0 0

subsection Cembrae

cembra^'^ 90-270 100

s/b/r/ca" 60- 90 3

pumila^ 120-150 3

koraiensis^'^ 90 3

'In USDA FS 1974.

^Characterized as exhibiting "extreme dormancy" (requiring more than 60
days of stratification in USDA FS 1974).

'Characterized as "suspected of having immature embryos at the time of
collection" in USDA FS 1974.

is realistic. Our intention in speculating is to stimulate
fiirther study of this possibility. All of this illustrates that
to understand whitebark pine's variable seed dormancy,
the nutcracker too must be considered. As pointed out by
Lanner (1980): "The ecology and evolution of the wingless-
seeded 'Bird Pines' can only be understood in a frame of
reference that includes the behavior of their feathered dis-
persers." The bare facts of the mutualism of whitebark
pine and Clark's nutcracker are by now well established.
The time is ripe to study the complexities of the relation-
ship, and its second-order effects.

REFERENCES

Balda, R. P. 1992. [Conversation with Ronald M. Lanner].
August 14. Flagstaff, AZ: Northern Arizona University.

Bialobok, S., ed. 1975. Stone-pine (Pinus cembra L.). Trans-
lated from Polish for U.S. Department of Agriculture and
National Science Foundation. National Technicsd Infor-
mation Service TT-73-54045. 173 p.

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis-
tribution of the pines of the world. Misc. Publ. 991.

210

Washington, DC: U.S. Department of Agriciilture, Forest
Service. 97 p.

Farris, G. J. 1983. California pignolia: seeds of Pinus sabini-
ana. Economic Botany. 37: 201-206.

Gifford, D. J. 1988. An electrophoretic analysis of the seed
proteins from Pinus monticola and eight other species of
pine. Canadian Journal of Botany. 66: 1808-1812.

Hubbard, W. D. 1977. Comparison of various methods for
the extraction of total lipids, fatty acids, cholesterol, and
other sterols from food products. Journal of the American
Oil Chemists' Society. 54: 81-83.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and estabhshment
of whitebark pine. Oecologia. 55: 192-201.

Kendall, K 1983. Use of pine nuts by black and grizzly bears
in the Yellowstone area. In: Meslaw, Charles E., ed. Bears —
their biology and management. West Glacier, MT: Inter-
national Association for Bear Research and Management:
166-173.

Kozlowski, T. T.; Gunn, C. R. 1972. Importance and charac-
teristics of seeds. In: Kozlowski, T. T., ed. Seed biology.
New York: Academic Press: 1-20.

Lammer, D. L.; Gifford, D. J. 1989. Lodgepole pine seed
germination. II. The seed proteins and their mobihzation
in the megagametophyte and embryonic axis. Csmadian
Journal of Botany. 67: 2544-2551.

Lanner, R. M. 1980. Avisin seed dispersal as a factor in the
ecology and evolution of limber and whitebark pines.
In: Dancik, B. P.; Higginbotham, K O., eds. Sixth North
American forest biology workshop proceedings; 1980
August 11-13; Edmonton, AB. Edmonton, AB: University
of Alberta: 15-45.

Lanner, R. M. 1981. The pinon pine: a natural and cultural
history. Reno, NV: University of Nevada Press. 208 p.

Lanner, R. M. 1982. Adaptations of whitebark pine for seed
dispersal by Clark's nutcracker. Canadian Journal of
Forest Research. 12: 391-402.

Mattson, D. J.; Jonkel, C. 1990. Stone pines and bears. In:
Schmidt, W. C; McDonald, K. J., comps. Proceedings —
sjnnposium on whitebark pine ecosystems: ecology of a
high-mountain resoiu-ce; 1989 March 29-31; Bozeman,
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart-
ment of Agriculture, Forest Service, Intermoimtain Re-
search Station: 223-236.

McCarthy, M. A.; Matthews, R. H. 1984. Composition
of foods: nut and seed products. Agric. Handb. 8-12.
Washington, DC: U.S. Department of Agriculture, Hu-
man Nutrition Information Service. 137 p.

McCaughey, W. 1992. [Letter to Ronald M. Lanner].
February 25. 1 leaf. Bozeman, MT: U.S. Department

of Agriculture, Forest Service, Intermountain Research
Station.

Parrish, J. W., Jr.; Martin, E. W. 1977. The effect of dietary
lysine level on the energy and nitrogen balance of the
dark-eyed junco. Condor. 79: 24-30.

Pitel, J. A.; Wang, B. S. P. 1990. Physical and chemical
treatments to improve germination of whitebark pine
seeds. In: Schmidt, W. C; McDonald, K J., comps.
Proceedings — sjmiposiiun on whitebark pine ecosystems:
ecology of a high-moxmtain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 130-133.

Schirone, B.; Piovesan, B.; Bellarosa, R.; Pelosi, C. 1991. A
teixonomic analysis of seed proteins in Pinus spp. (Pina-
ceae). Plant Systematics and Evolution. 178: 43-53.

Shimanyuk, A. P., ed. 1963. Fruiting of the Siberian stone
pine in east Siberia. Translated from Russian for U.S.
Department of Agricultural and National Science Foim-
dation by Israel Program for Scientific Translations.
Jerusalem. 204 p.

Tikhomirov, B. A. 1939. Khozyaistvenoye ispol'zovaniye
kedrovogo stlannika. Sov. Sever. 4: 62-65.

Tomback, D. F. 1978. Foraging strategies of Clark's nut-
cracker. Living Bird. 16: 123-161.

Turcek, F. J.; Kelso, L. 1968. Ecological aspects of food
transportation and storage in the Corvidae. Communica-
tions in Behavioral Biology Part A. 1: 277-297.

U.S. Department of Agriculture, Forest Service. 1974. Seeds
of woody plants in the United States. Agric. Handb. 450.
Washington, DC: U.S. Department of Agricultiire, Forest
Service. 883 p.

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of the
Clark's nutcracker and the pinon pine for efficient seed
harvest and dispersal. Ecological Monographs. 47: 89-111.

Vander Wall, S. B.; Hutchins, H. E. 1983. Dependence of
Clark's nutcracker, Nucifraga columbiana, on conifer
seeds dming the postfledging period. Canadian Field
Naturahst. 97: 208-214.

Van Etten, C. H.; Kwolek, W. F.; Peters, J. E.; Barclay, A. S.
1967. Plant seeds as protein sources for food or feed. Eval-
uation based on amino acid composition of 379 species.
Journal of Agricultural and Food Chemistry. 15: 1077-1089.

White, A.; Handler, P.; Smith, E. L.; Stetten, D. 1959. Prin-
ciples of biochemistry. New York: McGraw-Hill Book Co.
1149 p.

Yoon, T.-H.; Im, K-J.; Koh, E. T.; Ju, J.-S. 1989. Fatty acid
compositions of Pinus koraiensis seed. Nutrition Re-
search. 9: 357-361.

211

BEAR USE OF WHITEBARK PINE
SEEDS IN NORTH AMERICA

David J. Mattson
Daniel P. Reinhart

Abstract — Whitebark pine (Pinus albicaulis) seeds are an im-
portant high-quality food for bear populations that occupy ecosys-
tems with continental climates south of the United States-
Canada border. Availability of pine seeds affects levels of
human-bear conflict and bear mortality. In most areas bears
acquire whitebark pine seeds by excavating red squirrel (Tamia-
sciurus hudsonicus) food caches. Squirrel densities appear to
limit bear use of pine seeds more than abundance of whitebark
pine. Thus management of whitebark pine habitats for bear use
primarily involves management to favor red squirrels. Bear use
of pine seeds is restricted to stands >100 years old, and can per-
sist for an additional 200 to 300 years. Consequently stand rota-
tions of 300 years and harvest rates of 3 percent per decade are
recommended. Also, because of potential or demonstrated nega-
tive impacts associated with global climate change and white
pine blister rust {Cronartium ribicola), long-term prospects for
whitebark pine should be part of considerations when managing
habitats of bear populations, especially insular populations.

Whitebark pine seeds are a predictably high-quality
bear food given their relatively leirge size, high fat con-
tent, and moderately high digestibility (table 1). Fat con-
tent is perhaps the most important attribute of these
seeds, given the importance of body fat to the life strategy
of bears (Stirling and Derocher 1990) and the greater effi-
ciencies of body fat accimiulation associated with high di-
etary fat content (Hadley 1985; Schemmel 1976). On the
other hand, use of whitebark pine seeds especially by griz-
zly bears (Ursus arctos horribilis) is complicated by the in-
dehiscent nature of whitebark pine cones (Lanner 1990),
and the consequent reliance of most bears on vdndthrow
or rodent intermediaries for access to the seeds (Kendell
1983; Mattson and Jonkel 1990; Taylor 1964).

Black (Ursus americanus) and grizzly bears have prob-
ably made use of pine seeds for as long as these two spe-
cies have coexisted with whitebark pine, approximately
2 million years for black bears and 12,000 years for griz-
zly bears (Axelrod 1986; Km-ten and Anderson 1980). Al-
though aboriginal Americans were probably aware of this
relationship, the first written records of bears using
whitebark pine seeds did not appear until well into the

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

David J. Mattson is Wildlife Biologist, Interagency Grizzly Bear Study
Team, Forestry Sciences Laborestry, Montana State University, Bozeman,
MT 59717; Daniel P. Reinhart is Management Biologist, National Park
Service, Resource Management, Yellowstone National Park, WY 82190.

20th century, in articles by authors such as Cahalane
(1947), Taylor (1964), and Tisch (1961). Details of the re-
lationship between bears suid whitebark pine seeds have
only recently been published, focusing on interior areas
characterized by more continental climates (Craighead
and others 1982; Kendall 1983; Mattson and Jonkel 1990;
Mattson and others 1992, 1993). In this paper we update
information presented by Mattson and Jonkel (1990), and
further elaborate on some aspects of the relationship be-
tween bears and whitebark pine in North America.

PATTERNS OF BEAR USE
Geographic Distribution

Bears consistently make use of whitebark pine seeds
only south of the United States-Ceinada border, where
whitebark pine is a potentially abundant component of
subalpine forests (fig. 1). However, even in this eirea ma-
jor differences in intensity of bear use are evident, associ-
ated with maritime and continental climates.

We examined published bear food habits studies fi-om
western North America that included the results of 13,130
bear feces examinations. Our analysis clearly showed
that the consistency and relative frequency of whitebark
pine seed use was much greater (18.7 percent) in areas

Table 1— Size, nutrient content, and relative digestibility by ursids
of whitebark pine seeds; compiled from Craighead and
others (1982), Kendall (1983), Lanner and Gilbert (these
proceedings), Mealey (1975), Pritchard and Bobbins
(1990), Weaver and Forcella (1986), and unpublished
Interagency Grizzly Bear Study Team data

Mean

Range of

Range

among

standard

Parameters

of means

studies

deviations

Seeds per cone

75-88

26-28

Dry seed weight (g)

0.07-0.10

0.09

0.01-0.02

Percent protein content^

11.9-14.2

12.8

Percent ether extract^

21 .8-30.4

27.1

Percent nitrogen-free extract'

12.4-27.2

19.8

Percent crude fiber'

34.8

(34.8)

Percent total dietary fiber' ^

40.3

(40.3)

Percent digestible energy

35O.I/M8.7

49.4

31.4

'Based on whole seed dry weight.
^Calculated for pinyon pine (Pinus edulis) seeds.
^Calculated based on laboratory feeding trials (Pritchard and Robbins
1990).

^Calculated from differential between field samples of fresh whole material
and feces (Mealey 1975).

212

Figure 1 — Distribution of whitebark pine and use of whitebark pine seeds by bears
in North America. Sparser Canadian distribution of whitebark pine is indicated by
dotted lines. Three different levels of pine seed use (<1 percent, 1-10 percent, and
>10 percent frequency in bear feces) are also indicated.

213

with continental climates compared to areas with mari-
time climates (2.7 percent) or areas in Canada where
whitebark pine was sparse (2.8 percent) (df = 2, G = 987.3,
P< 0.001).

These differences are logically attributable to several
factors. In areas where whitebark pine is scarce, the
bears may not consider it to be a food, or not choose to use
its seeds if they are disjunct from other higher value feed-
ing opportunities. Areas influenced by maritime climates
are characterized by an abundance of berry-producing
shrubs, typically at elevations much lower than whitebark
pine (Jonkel and Cowan 1971; Poelker and Hartwell 1973;
Servheen 1983; Tisch 1961; and others). These berries
are probably more valuable than whitebark pine seeds,
and so local bears would logically tend to use berry-rich
habitats to the exclusion of whitebark pine habitats (Tisch
1961). Conversely, in areas with continental climates ber-
ries are typically fewer and whitebark pine more abun-
dant (Aune cuid Kasworm 1989; Mattson and others
1991). We hypothesize that for this reason bears would
logically focus more on the consumption of whitebark pine
seeds. Recent heavy whitebark pine mortality in mari-
time areas, caused by white pine blister rust, has also ar-
guably had negative effects on bear use of pine seeds in
these areas (Kendall and Arno 1990; Mattson and Jonkel
1990).

Because the large majority of whitebark pine seed use
by bears has occurred in interior ecosystems, our subse-
quent discussion will focus on these areas; and because
most research concerning bear use of pine seeds comes
from the Yellowstone ecosystem, most of the results pre-
sented here are from this study area.

APR MAY JUN JUL AUG SEP OCT
n = 5 10 14 14 14 11 6

Figure 2 — Mean percent volume of pine
seeds in Yellowstone grizzly bear feces by
month, 1977-90. Maximum recorded volume
is indicated by dashed lines and mean per-
cent volume of pine seeds in feces of occur-
rence is shown in the inset.

Temporal Variation in Use

Seasonal Variation — Because whitebark pine seeds
are durable, bears use not only seeds contemporary to
their maturation, but also seeds that survive for up to
1 year after initial availability of a large crop (Kendall
1983; Mattson and Jonkel 1990; Mattson and others 1993).
Current-year use usually begins after mid-August with
onset of caching by red squirrels (Hutchins and Lanner
1982), but can vary from late July to early September,
apparently depending on seed maturation (Mattson and
others 1993). Use of overwintered seeds does not usually
begin imtil after mid-May (Mattson and Jonkel 1990), but
can begin as early as the first week of April and involve
excavation of cones out of squirrel middens, through snow
0.3-1.2 m deep (Kendall 1983).

Use of both current-year's and overwintered seeds by
bears results in a primary September-October peak and
a secondary June peak in frequency of pine seeds in bear
feces, when averaged over a number of years (fig. 2).
With the exception of July, bear feces tend to consist
almost exclusively of pine seed remains (>80 percent)
when bears are in the whitebark pine zone feeding on
pine seeds (fig. 2).

Annual Variation — Use of pine seeds by bears varies
considerably among years. In the Yellowstone area, pine
seeds have dominated the diet during some years (for ex-
ample 1979) and been virtually absent other years (for
example 1977, 1982, and 1988) (Mattson and others 1991,
1992). This variation in use is related to seed crop size,
with heavy use occurring only when crops average > 13-23
cones per tree (from counts on permanent transects), ap-
proaching an asymptote of use (75 percent frequency in
feces) at crop sizes averaging 38 cones per tree. Use of
overwintered pine seeds during the following June-July is
similarly related to crop size, but with inflection points of
20-29 seeds and an asymptote of 38 percent fi-equency in
feces (fig. 3). This relationship between bear use and crop
size is acutely sigmoidal, suggesting the existence of a use
threshold that is possibly related to sattiration of demand
by competing birds and rodents, nonlinear increases in
foraging efficiency, or both. The higher thresholds and
lower asymptote associated with use of overwintered
seeds is attributable to depletion of the crop diiring the
previous fall and winter.

Varying inflection points in the hypothesized relation-
ship between bear use and seed crop size reflect an appar-
ent trend over the years 1980-91 toward lower thresholds
to substantial bear use (fig. 3). The years 1980 and 1981
are further distinguished by the relative insensitivity of
bear use to crop size compared to later years. Interest-
ingly, asymptotes associated with maximimi potential
bear use have not varied significantly diiring this same
time period. At the very least, this variance suggests
that the relative intensity with which bears use pine
seeds in a given study area can change. The causes of
these apparent changes can only be speculated, including:
increasingly efficient use of comparable-sized crops by the
Yellowstone bear population; and an increasing willing-
ness to use seeds imder less optimal conditions, attribut-
able either to increasing bear densities or decreasing over-
all carrying capacity.

214

AUGUST-OCTOBER USE, CONTEMPORARY YEAR

JUNE-JULY USE, FOLLOWING YEAR

100

25 —

-[((CONES/60)-1)/I]'^ r^ = 0 99
% FREQ = 38.0 * e f = 353 3

10 20 30 40

MEAN NUMBER OF CONES PER TREE

Figure 3 — A, relationship between frequency
of whitebark pine seeds in August-October
Yellowstone grizzly bear feces and mean num-
ber of cones counted per tree on permanent
transects during the contemporary year, 1980-
91 . B, relationship between frequency of pine
seeds in June-July feces and mean number of
cones counted the previous fall, 1 980-91 .

Interestingly, in a part of Montana characterized by
a more maritime climate and relatively little bear use of
pine seeds, Tisch (1961) counted an average of five and
16 cones per tree during 1959 and 1960, respectively.
Both of these averages were less than the approximately
20 cones per tree associated with the onset of heavy pine
seed use in the Yellowstone area.

Habitats Associated With Use

Elevation and Aspect — As would be expected, bear
use of whitebark pine seeds is restricted to elevations
above the lower limits of whitebark pine distribution.
This lower elevational limit varies with latitude and
longitude, from about 2,400 m in the Yellowstone area

(44.5°N.). to 1,890 m along the East Front of the Montana
Rockies (47.5° N.) (Mattson and Jonkel 1990).

However, within this constraint bear use varies consid-
erably among aspects and elevations, depending on the
year. We found that diiring three successive large seed
crops in the Yellowstone area, bear use was nonrepetitive
among broad aspectual and elevational categories; during
1985-86 heaviest on east aspects at mid-elevations, during
1987 heaviest on west and north aspects at low- to mid-
elevations, and dxuing 1989-90 heaviest on east and south
aspects at high elevations (Mattson and others 1993). We
also found differences in elevational distribution of pine
seed use between two successive crops at a local scale,
in the Cooke City area northeast of Yellowstone National
Park, with use concentrated at elevations >2,870 m
dxoring 1990 and at elevations <2,720 m during 1991
(Reinhart and Mattson 1992). This geographic variation
was inconclusively related to seed production, but demon-
strated the potential for considerable differences in habi-
tat distribution of pine seed use over a 6- to 7-year period.

Habitat Tjrpes — Ecosystems in the Western United
States have been described in terms of ecological land-
scape units called habitat types, that are further aggre-
gated as series and split into phases (Pfister and others
1977; Steele and others 1983). In the Yellowstone area
most bear use of pine seeds has occurred with varying in-
tensity in 14 different habitat types. In accord with varia-
tion in distribution of pine seed use among aspects and el-
evations, distribution of pine seed use by bears has also
varied among habitat types (Mattson and others 1993).
Even so, the Abies lasiocarpa /Vaccinium scoparium-
Pinus albicaulis phase, and the Abies lasiocarpa /
Calamagrostis canadensis and Abies lasiocarpa I
Thalictrum occidentale habitat types have been used
intensively during all years when pine seeds are abtm-
dant (Mattson and Jonkel 1990). Conversely, use of habi-
tat types at lowest and highest elevations of the white-
bark pine zone has been most varied. Habitats in interior
ecosystems where whitebark pine is a chmax overstory
species have been only rarely used by bears for foraging
on pine seeds (Mattson and Jonkel 1990). However, even
in these habitats heavy use can occur, as diiring 1989-90
in the Yellowstone area (Mattson and others 1993).

Distribution of pine seed use among habitat types
was surprisingly consistent between study areas in the
Yellowstone ecosystem and along the East Front of the
Montana Rocky Mountains (Mattson and Jonkel 1990).
This suggests that patterns of habitat use associated with
bear consumption of pine seeds are consistent among
areas vvdth continental climates, and that more detailed
results pertaining to habitat use from the Yellowstone
ecosystem are extrapolable over interior regions of the
United States. Because very Uttle information is avail-
able concerning bear use of pine seeds in more maritime
ecosystems, we can make no inferences about patterns of
use in these areas other than it being restricted to sites
with mature cone-producing whitebark pine.

Red Squirrel and Whitebark Pine Abundance —

Red squirrel density is probably the most important proxi-
mal factor affecting bear use of pine seeds in interior eco-
systems. The importance of red squirrels to especially

215

grizzly bear foraging has been observed by Kendall (1983),
Mattson and Jonkel (1990), and Schallenberg and Jonkel
(1980). Craighead and others (1982), Graber and White
(1983), Murie (1944), and Raine and Kansas (1990) also
mentioned rodent or squirrel caches as potential sources
of pine seeds for bears, but were more equivocal about the
overall importance of these intermediaries. Interestingly,
neither Craighead and Craighead (1972) nor Mealey
(1975) mention that grizzly bears raided red squirrel food
caches during their studies in Yellowstone. The disparity
between these and later observations in the Yellowstone
area is not easily reconciled, and suggests that differences
could be at least partly attributable to variation in bear
behavior and pine seed availability described earlier.

Most (86 percent) of 232 sites where grizzly bears were
known to have used pine seeds in the Yellowstone area,
1986-91, involved bears excavating red squirrel cone
caches. All but five of the other 32 instances where griz-
zly acquired pine seeds by scavenging fallen cones oc-
curred during 1989-90 when much of the cone production
occurred at elevations above the normal range of red
squirrels (Mattson and others 1993; Reinhart and
Mattson 1990). Conversely, 51 percent of 53 sites where
black and grizzly bears used pine seeds near Cooke City,
MT, during 1990-91 were characterized by scavenging of
fallen cones (Reinhart and Mattson 1992). This observa-
tion suggests that scavenging of fallen cones can be locally
important. However, all 26 instances when bears scav-
enged fallen cones in this area occurred during 1990, in
conjunction with use of seeds overwintered from the im-
usually high-elevation crop of 1989.

Scavenging apparently predominates over excavation
of squirrel caches in environments where squirrels are
rare. Red squirrel densities in the whitebark pine zone
are strongly related to a site favorability index (r^ = 0.79),
with squirrel densities increasing as exposm*e to prevail-
ing winds and elevation decreases (Reinhart and Mattson
1990). Given that basal area and exclusivity of whitebark
pine increase with wind exposure and elevation (Mattson
and Reinhart 1990), there is an inherently negative rela-
tionship between dominance of whitebark pine and den-
sity of red squirrels within the whitebark pine zone. Thus,
scavenging predominates over midden excavation where
whitebark pine comprises >75 percent of the overstory
(four categories of whitebark pine percentage; df = 3,
G = 69.7, P < 0.001), and at elevations >2,870 m (four
elevation categories; df = 3, G = 42.4, P < 0.001); where
red squirrels are rare.

Bear use of pine seeds is clearly the result of opportuni-
ties implicit to the varying abundance of red squirrels and
whitebark pine. However, there is reason to hypothesize
that squirrel densities impose greater constraints on bear
use of pine seeds than does abundance of whitebark pine.
We analyzed data from 1987 to examine this hypothesis,
looking at the relationship between unit area volume of
excavations for pine seeds by Yellowstone grizzly bears
and red squirrel density and whitebark pine basal area
(fig. 4). At any given squirrel density, bear use abruptly
increased and approached maximum levels at relatively
low basal areas of whitebark pine, between approximately
3 and 8 m^/ha. Conversely, the relationship between bear
activity and red squirrel densities was more nearly linear.

BEAR USE = 22.6 • (e
^Opr

< I

-[((SD/3.68)-1)/0.382]'

)*(e

-[((BA/254)-1)/0.983l*

<
ui

QC
<

<
m

bi.
a.
<
m

LU
X

1.0 2.0 3.0 ^.0

SQUIRREL DENSITY (N/KM) (SD)

Figure 3 — The relationship between intensity of
grizzly bear feeding on pine seeds, as indicated by
density of excavations (m^/km) and squirrel density
and whitebark pine basal area, Yellowstone area,
1987.

with excavations gradually increasing to a maximum that
corresponded to maximum squirrel densities observed in
the whitebark pine zone. Basically, whitebark pine basal
area seemed to be associated primarily with the probabil-
ity that a given midden would be excavated, sometimes
repeatedly, by bears (Mattson and Reinhart 1987), with
the greatest effects at very low whitebark pine abun-
dance. Once this low threshold of whitebark pine abun-
dance was exceeded (> about 4.3 m^/ha), red squirrel den-
sities appeared to limit bear use of pine seeds during most
years.

Timber Stand Age — Bear use of whitebark pine seeds
in the Yellowstone area has occurred almost exclusively in
mature stands. According to a successional classification
of timber types by Despain (1990), all bear use of pine
seeds from squirrel middens and most bear use of scav-
enged seeds has occurred in "mature" cover types, typi-
cally designated LP2, LP3, SF, WB2, WB3, WB, or DF3.
Results from a use-availability analysis of cover types in
Yellowstone National Park corroborate the affinity of griz-
zly bears for later successional cover types when using
pine seeds (Knight and others 1984). This analysis showed
that the early successional WBl cover type was consis-
tently imder-used or used without preference during all
seasons, whereas mature WB cover types were consistently
used with disproportionally greater intensity than ex-
pected, especially dining the fall, primarily for the con-
sumption of pine seeds.

216

Based on ages given by Despain (1990) cover types asso-
ciated with pine seed use are rarely less than 100 years
old. Ages of 160 stands in two intensive Yellowstone
study areas located in the whitebark pine zone (Mattson
and Reinhart 1987) also revealed that none of these ma-
ture cover types was <125 years old.

These minimum stand ages correspond with use ex-
pected by levels of cone production. Weaver and others
(1990) observed that whitebark pine cone production was
relatively low in stands <100 years old and reached peak
levels soon thereafter. Morgan and Biinting (in press)
also observed that younger WBO and WBl stands pro-
duced few cones and had few excellent crops. Both of
these authors observed that cone production persisted at
least at moderate levels in much older stands, up to 300 to
400 years old. These observations correspond well with
the resvilts of more exhaustive studies of the closely re-
lated Siberian stone pine (Pinus sibirica) (Axelrod 1986)
by Russian researchers (Mattson and Jonkel 1990).

These relationships of use to stand age have several im-
plications. First, potential productivity of whitebark pine
sites, at least for bears, is not realized until after approxi-
mately 100 years, and can persist for 200 to 300 years
thereafter. Thus, in contrast to many other temperate
ecosystems where production of bear foods is concentrated
on berry-producing shrubs in early successional stages,
production of bear foods in the whitebark pine zone is con-
centrated in much older mid- to late-successional stands
(Mattson 1990). Second, given these time frames, long-
term productivity of the whitebark pine zone in interior
ecosystems would be best assiired by disturbance regimes
that replaced approximately 3 percent of mature acreage
per decade.

EFFECTS ON BEAR DEMOGRAPHY
Fecundity

Whitebark pine seeds are an important so\irce of energy
for bears that live in interior ecosystems of the United
States. Craighead and others (1982) estimated that pine
seeds were the most important soxirce of energy for grizzly
bears in their study area located east of the Continental
Divide in the Rocky Mountains of Montana. Preliminary
estimates by Servheen and others (1986) suggested that
Yellowstone grizzly bears derived about 23 percent of
their net digested energy from whitebark pine seeds,
second only to the approximately 30 percent that they
derived from ungulate meat. In addition, preliminary
analysis of Yellowstone data suggests that female grizzly
bears make proportionately much greater use of pine
seeds compared to males. Thus there is a good basis for
hypothesizing that availability and use of pine seeds
affects the fec;mdity of female bears residing in these
ecosystems.

Further preliminary analyses of data from the Yellow-
stone ecosystem offer support for this hypothesis. Female
grizzly bears that frequently used whitebark pine seeds
appeared to exhibit substantially higher reproductive
rates and reproduce at a significantly earlier age than
females who consumed few pine seeds. Much of the differ-
ence in reproductive rate, as measured by both production

and survival of offspring to weaning, was attributable to
low survivorship of cubs dependent on dams that made
httle lose of pine seeds.

Mortality and Movements

Whitebark pine seed availability has even greater ef-
fects on svirvivorship of subadult males and adult females
than on fecimdity of reproducing females. During years
when pine seeds are used very little, conflicts with hu-
mans escalate dramatically in the Yellowstone area.
Management-related trappings of grizzly bears instituted
to alleviate these conflicts are, on average, six times as
nimierous during nonuse compared to use years (Mattson
and others 1992). As a direct result of this increased con-
flict, an average of two times as many adult females and
three times as many subadult males die during years
when pine seeds are not available (Mattson and others
1992).

This increased mortality is not attributable to increased
movement of htunans smiong bears or to starvation, but
rather to increased activity of bears around hvunan facili-
ties (Mattson and others 1992). Whitebark pine in the
Yellowstone and Rocky Movmtain East Front ecosystems
tends to occm* in areas remote from human facilities (Aune
and Kasworm 1989; Mattson and others 1992). Thus,
when bears are intensively using pine seeds they are typi-
cally far removed from himaans and hmnan-bear conflict
is correspondingly low. Conversely, when pine seeds are
not available bears seek out alternate foods that are often
closely associated with himian facilities. Historically
these alternate foods were often of human origin (Knight
and others 1988), but more recently these alternate foods
have been natives that tend to occupy lower elevation
habitats also occupied by humans (Mattson and Knight
1989). In addition, the greater movements of Yellowstone
grizzly bears diiring years or seasons when pine seeds are
not available (Blanchard and Knight 1991; Haroldson and
Mattson 1985) very likely exacerbate this unfavorable
situation and increase conflicts with himaans.

MANAGEMENT IMPLICATIONS

Because whitebark pine seeds are apparently so impor-
tant to bears in interior ecosystems of the United States,
whitebark pine deserves priority in the management of
bear habitat. As indicated earlier, both temporal and geo-
graphic variation in the availability of whitebark pine
seeds can have major implications on the survival and fe-
cimdity of bears in these ecosystems. Thus, long-term
variations in availability of pine seeds attributable to the
productivity and abundance of whitebark pines warrants
the concern and attention of managers.

As pointed out previously, stands with relatively little
whitebark pine can be of considerable value to bears as
long as they contain high densities of red squirrels. Thus,
management of bear habitat for availability of pine seeds
implies managment to meet red squirrel habitat require-
ments. In the whitebark pine zone this means simulta-
neously maximizing stand basal area, overstory species
diversity, and components of whitebark pine and Douglas-
fir (Pseudotsuga menziesii) (Reinhart and Mattson 1990).

217

This kind of favorable condition often coincides with com-
mercial forests occupying lower elevations of the white-
bark pine zone; and maintenance of these conditions po-
tentially conflicts with other management objectives
related to the extraction of wood products. In areas where
bear management is a high priority, timber harvest cycles
should therefore employ roughly 300-year stand rotations
in the whitebark pine zone, with landscape-wide harvest
levels at approximately 3 percent per decade.

Global climate change has potentially major implica-
tions to the abundance of whitebark pine in interior eco-
systems and consequently to the prospects of associated
bear populations. Romme and Turner (1991) projected
the fates of major life zones in the Yellowstone area under
three different climatic scenarios likely to be associated
with doubled content in the atmosphere. Under all
three scenarios they projected a decline of the whitebark
pine zone to <10 percent of its current extent. Without
wildfires, loss of whitebark pine may be prolonged due to
the long lifespan of the species. However, increasingly
frequent fires, especially under a drier scenario, could
hasten the demise of whitebark pine by eliminating it
from sites where it coiild not reestablish (Romme and
Turner 1991). Thus, with global climate warming a real
possibility, managers should be cautious about harvesting
any whitebark pine, especially at lower elevations. In ad-
dition, judicious control of natural fires may be desirable
to favor the persistence of existing mature whitebark
pine.

The potential and demonstrated negative impacts of
global climate change and white pine blister rust on an
important source of bear food like whitebark pine point up
the legitimate need for concern about long-term changes
in bear habitats. This consideration is especially relevant
to the management of insular bear populations, exempli-
fied by Yellowstone's grizzly bears (Mattson and Reid
1991). In situations where occupied bear habitat is not
only isolated but also restricted, bear populations are es-
pecially vulnerable to changes in carrying capacity be-
cause of limited options to shift or expand their ranges or
benefit from the infliix of bears from other nearby or con-
tiguous populations. Thus, uncertainty over the long-
term prospects of whitebark pine should be a major con-
sideration in management decisions, especially those that
might otherwise reduce habitat capability under the as-
sumption of long-term habitat stability.

Finally, managers should be concerned about the place-
ment of additional human facilities anywhere in the
whitebark pine zone. Currently, most of the whitebark
pine zone serves as a refuge for bears from conflict with
humans, at least when pine seeds are available. This ref-
uge effect will be diminished with every additional human
intrusion. Wherever bear management is a priority, addi-
tional human facilities should not be constructed in the
whitebark pine zone and existing facilities should be re-
moved if at all possible.

REFERENCES

Aune, Keith; Kasworm, Wayne. 1989. Final report East
Front grizzly studies. Helena, MT: Montana Depart-
ment of Fish, Wildlife, and Parks. 332 p.

Axelrod, Daniel I. 1986. Cenozoic history of some western
American pines. Annals of the Missouri Botanical Gar-
dens. 73: 565-641.

Blanchard, Bonnie M.; Knight, Richard R. 1991. Move-
ments of Yellowstone grizzly bears, 1975-87. Biological
Conservation. 58: 41-67.

Cahalane, Victor H. 1947. Mammals of North America.
New York: Macmillan. 682 p.

Craighead, Frank C, Jr.; Craighead, John J. 1972. Griz-
zly bear prehibernation and denning activities as deter-
mined by radio-tracking. Wildlife Monograph 32. 35 p.

Craighead, John J.; Sumner, J. S.; Scaggs, G. B. 1982. A
definitive system for analysis of grizzly bear habitat
and other wilderness resources. Wildlife-Wildlands
Institute Monograph 1. Missoula, MT: University of
Montana, U of M Foundation. 279 p.

Despain, Donald G. 1990. Yellowstone vegetation,
Boiilder, CO: Roberts Reinhart. 239 p.

Graber, David M.; White, Marshall. 1983. Black bear food
habits in Yosemite National Park. In: International
Conference on Bear Research and Management: Pro-
ceedings: 5: 1-10.

Hadley, N. F. 1985. The adaptive role of lipids in biologi-
cal systems. New York: John Wiley and Sons. 319 p.

Haroldson, Mark; Mattson, David J. 1985. Response of
grizzly bears to backcountry himian use in Yellowstone
National Park. Bozeman, MT: U.S. Department of the
Interior, National Park Service, Interagency Grizzly
Bear Study Team. 38 p.

Jonkel, Charles J.; Cowan, Ian McT. 1971. The black bear
in the spruce-fir forest of northwestern Montana. Wild-
life Monograph. 27. 55 p.

Kendall, Katherine C. 1983. Use of pine nuts by black and
grizzly bears in the Yellowstone area. In: International
Conference on Bear Research and Management: Pro-
ceedings: 5: 166-173.

Kendall, Katherine C; Arno, Stephen F. 1990. Whitebark
pine — an important but endangered wildlife resource.
In: Schmidt, Wyman C; McDonald, Kathy J., comps.
Proceedings — symposium on whitebark pine ecosys-
tems: ecology and management of a high-mountain re-
source; 1989 March 29-31; Bozeman, MT. Gen. Tech.
Rep. INT-270. Ogden, UT: U.S. Department of Agricul-
ture, Forest Service, Intermountain Research Station:
264-273.

Knight, R. R.; Blanchard, B. M.; Eberhardt, L. L. 1988.
Mortality patterns and population sinks for Yellowstone
grizzly bears, 1973-1985. Wildlife Society Bulletin. 16:
121-125.

Knight, Richard R.; Mattson, David J.; Blanchard,
Bonnie M. 1984. Movements and habitat use of the
Yellowstone grizzly bear. Bozeman, MT: U.S. Depart-
ment of the Interior, National Park Service, Interagency
Grizzly Bear Study Team. 177 p.

Kurten, Bjorn; Anderson, Elaine. 1980. Pleistocene mam-
mals of North America. New York: Columbia University
Press. 512 p.

Lanner, Ronald M. 1990. Biology, taxonomy, evolution,
and geography of stone pines of the world. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposium on whitebark pine ecosystems: ecology and
management of a high-mountain resource; 1989 March

218

29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden,
UT: U.S. Department of Agriculture, Forest Service,
Intermountain Research Station: 14-24.

Mattson, David J. 1990. Human impacts on bear habitat
use. In: International Conference on Bear Research and
Management: Proceedings: 8: 33-56.

Mattson, David J.; Blanchard, Bonnie M.; Knight, Richard
R. 1991. Food habits of Yellowstone grizzly bears, 1977-
1987. Canadian Journal of Zoology. 69: 1619-1629.

Mattson, David J.; Blanchard, Bonnie M.; Knight, Richard
R. 1992. Yellowstone grizzly bear mortality, hvmaan ha-
bituation, and whitebark pine seed crops. Journal of
Wildlife Management. 56(3): 432-442.

Mattson, David J.; Jonkel, Charles. 1990. Stone pines and
bears. In: Schmidt, Wyman C; McDonald, Kathy J.,
comps. Proceedings — symposium on whitebark pine
ecosystems: ecology and management of a high-moun-
tain resource; 1989 March 29-31; Bozeman, MT. Gen.
Tech. Rep. INT-270. Ogden, UT: U.S. Department of
Agriculture, Forest Service, Intermountain Research
Station: 223-236.

Mattson, David J.; Knight, Richard R. 1989. Evaluation of
grizzly bear habitat using habitat type and cover type
classifications. In: Ferguson, Dennis E.; Morgan,
Penelope; Johnson, Frederic D., comps. Proceedings —
land classifications based on vegetation: applications for
resource management; 1987 November 17-19; Moscow,
ID. Gen. Tech. Rep. INT-257. Ogden, UT: U.S. Depart-
ment of Agriculture, Forest Service, Intermountain
Research Station: 135-143.

Mattson, David J.; Reid, Matthew M. 1991. Conservation
of the Yellowstone grizzly bear. Conservation Biology.
5: 364-372.

Mattson, David J.; Reinhart, Daniel P. 1987. Grizzly bear,
red squirrels, and whitebark pine: third year progress
report. In: Knight, Richard R.; Blanchard, Bonnie M.;
Mattson, David J. Yellowstone grizzly bear investiga-
tions: report of the Interagency Study Team, 1986.
Bozeman, MT: U.S. Department of the Interior,
National Park Service, Interagency Grizzly Bear Study
Team: 36-45.

Mattson, David J.; Reinhart, Daniel P. 1990. Whitebark
pine on the Mount Washburn massiff, Yellowstone Na-
tional Park. In: Schmidt, Wyman C; McDonald, Kathy
J., comps. Proceedings — symposium on whitebark pine
ecosystems: ecology and management of a high-moun-
tain resource; 1989 March 29-31; Bozeman, MT. Gen.
Tech. Rep. INT-270. Ogden, UT: U.S. Department of
Agriculture, Forest Service, Intermountain Research
Station: 106-117.

Mattson, David J.; Reinhart, Daniel P.; Blanchard,
Bonnie M. 1993. Variation in production and bear use of
whitebark pine seeds in the Yellowstone area. In: Pro-
ceedings — conference on plants and their environments;
1991 September; Mammoth, WY. Denver, CO: U.S.
Department of the Interior, National Park Service,
Regional Office. [In press].

Mealey, Stephen Patrick. 1975. The natural food habits of
free ranging grizzly bears in Yellowstone National
Park, 1973-1974. Bozeman, MT: Montana State Univer-
sity. 158 p. Thesis.

Murie, O. J. 1944. Progress report on the Yellowstone
bear study. Mammoth, WY: U.S. Department of the
Interior, National Park Service, Yellowstone National
Park. 13 p.

Pfister, Robert D.; Kovalchik, Bernard L.; Arno, Stephen F.;
Presby, Richard C. 1977. Forest habitat types of Mon-
tana. Gen. Tech. Rep. INT-34. Ogden, UT: U.S. Depart-
ment of Agricultiu-e, Intermountain Forest and Range
Experiment Station. 174 p.

Poelker, Richard J.; Hartwell, Harry D. 1973. Black bear
of Washington. Biol. Bull. 14. Olympia, WA: Washing-
ton State Game Department. 180 p.

Pritchard, Geoffrey T.; Robbins, Charles T. 1990. Diges-
tive and metabolic efficiencies of grizzly and black
bears. Canadian Journal of Zoology. 68: 1645-1651.

Raine, R. M.; Kansas, J. L. 1990. Black bear seasonal food
habits and distribution by elevation in Banff" National
Park, Alberta. In: International Conference on Bear Re-
search and Management: Proceedings: 8: 297-304.

Ronmie, William H.; Turner, Monica G. 1991. Implica-
tions of global climate change for biogeographic pat-
terns in the greater Yellowstone ecosystem. Conserva-
tion Biology. 5(3): 373-386.

Reinhart, Daniel P.; Mattson, David J. 1990. Red squir-
rels in the whitebark zone. In: Schmidt, Wyman C;
McDonald, Kathy J., comps. Proceedings — symposium
on whitebark pine ecosystems: ecology and manage-
ment of a high-moimtain resource; 1989 March 29-31;
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT:
U.S. Department of Agriculture, Forest Service, Inter-
mountain Research Station: 256-263.

Reinhart, Daniel P.; Mattson, David J. 1992. Grizzly bear
and black bear habitat use in the Cooke City, Montana,
area 1990-1991. Bozeman, MT: U.S. Department of the
Interior, National Park Service, Interagency Grizzly
Bear Study Team. 30 p.

Schallenberger, A.; Jonkel, C. J. 1980. Rocky Mountain
East Front grizzly studies, 1979. Annual Report. Border
Grizzly Project Special Report 39. Missoula, MT: Border
Grizzly Project. 207 p.

Schemmel, R. 1976. Physiological considerations of lipid
storage and utilization. American Zoologist. 16:
661-670.

Servheen, Christopher. 1983. Grizzly bear food habits,
movements, and habitat selection in the Mission Moim-
tains, Montana. Journal of Wildlife Management. 47(4):
1026-1035.

Servheen, Christopher; Knight, Richard; Mattson, David;
Mealey, Steven; Strickland, Dale; Varley, John; Weaver,
John. 1986. Report to the IGBC on the availability of
foods for grizzly bears in the Yellowstone ecosystem.
Missoula, MT: U.S. Fish and Wildlife Service, Office of
the Grizzly Bear Recovery Coordinator. 21 p.

Steele, Robert; Cooper, Stephen V.; Ondov, David M.;
Roberts, David W.; Pfister, Robert D. 1983. Forest
habitat types of eastern Idaho-western Wyoming. Gen.
Tech. Rep. INT-144. Ogden, UT: U.S. Department of
Agrictdture, Forest Service, Intermountain Forest and
Range Experiment Station. 122 p.

Stirling, Ian; Derocher, Andrew E. 1990. Factors affecting
the evolution and behavioral ecology of the modern

219

bears. In: International Conference on Bear Research
and Management: Proceedings: 8: 189-204.

Taylor, R. A. 1964. Coliimbian ground squirrel and cam-
bium found in grizzly bear stomachs taken in the fall.
Journal of Mammalogy. 45(3): 476-477.

Tisch, Edward L. 1961. Seasonal food habits of the black
bear in the Whitefish Range of northwest Montana.
Missoula, MT: University of Montana. 108 p. Thesis.

Weaver, T.; Forcella, F. 1986. Cone production in Pmi/s
albicaulis forests. In: Shearer, Raymond C, comp.
Proceedings — conifer tree seed in the Inland Mountain
West symposium.; 1985 August 5-6; Missoula, MT. Gen.

Tech. Rep. INT-203. Ogden, UT: U.S. Department of
Agricidture, Forest Service, Intermountain Resesirch
Station: 68-76.
Weaver, T.; Forcella, F.; Dale, D. 1990. Stand develop-
ment in whitebark pine woodlands. In: Schmidt,
Wyman C; McDonald, Kathy J., comps. Proceedings —
symposiimi on whitebark pine ecosystems: ecology and
management of a high-moimtain resom-ce; 1989 March
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden,
UT: U.S. Department of Agriculture, Forest Service, In-
termountain Research Station: 151-155.

220

ECOLOGICAL RELATIONSHIP
BETWEEN CLARK'S NUTCRACKER
AND FOUR WINGLESS-SEED STROBUS
PINES OF WESTERN NORTH AMERICA

Diana F. Tomback

Abstract — The Clark's nutcracker {Nucifraga columhiana) is an
important seed disperser for at least four western North American
Strohus pines: whitebark (Pinus alhicaulis), limber (P. flexilis),
Colorado pinyon {P. edulis), and singleleaf pinyon (P. mono-
phylla). Nutcrackers share the role of pinyon pine seed disperser
with the pinyon jay (Gymnorhinus cyanocephalus) and may also
contribute to seed dispersal of southwestern white pine (P. strobi-
formis). All five species of pines have large, wingless seeds. In
whitebark pine and the pinyon pines, ripe seeds are retained in
cones.

In this paper I briefly review the ecological relationship
between Clark's nutcracker {Nucifraga columhiana) and
four pine species of the subgenus Strohus that depend on
the nutcracker for seed dispersal. A fifth Strohus pine,
southwestern white {Pinus strohiformis) may also be dis-
persed by Clark's nutcracker (see Benkman and others
1984), but the requisite studies have not yet been done.

This review is not intended to be exhaustive but rather
to sketch the basics of these interactions, which are simi-
lar to those between the the Eurasian nutcracker {N. caryo-
catactes) and five European and Asian Strohus pines.

GEOGRAPHIC DISTRffiUTIONS

The Clark's nutcracker ranges throughout much of
the montane regions of the Western United States (AOU
1983). The eastern limits are the Black Hills of South
Dakota and, irregularly, the pine forests of southwestern
South Dakota and northwestern Nebraska, and east of
the Rocky Mountain firont. To the south, the nutcracker
ranges into northern Baja California, and to the north
into Canada along the Coast Mountains and Rocky Moun-
tains (fig. 1).

The geographic boundaries of all four Strohus pines
known to be dispersed by Clark's nutcrackers fall v^dthin
the range of the nutcracker (Critchfield and Little 1966)
(fig. 1). Whitebark pine {Pinus alhicaulis, Subsection
Cembrae) is a subalpine-to-treeline species that occurs
fi'om the Sierra Nevada north through the coast ranges
and Northern Rocky Mountains into Alberta and British

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Diana F. Tomback is Associate Professor of Ecology, Department of
Biology and Center for Environmental Sciences, University of Colorado
at Denver, Denver, CO 80217-3364.

Columbia. Limber pine {Pinus flexilis, Subsection Strohi),
which occupies a wide elevational range, is distributed
from southern California across the Great Basin to the
Rocky Mountains and north into Alberta and British
Columbia. Both the singleleaf pinyon {Pinus monophylla)
and Colorado pinyon pine {Pinus edulis), which are also
dispersed by pinyon jays {Gymnorhinus cyanocephalus)
(see, for example, Ligon 1978), are lower treeline species
in the Subsection Cemhroides. The singleleaf pinyon pine
ranges fi'om eastern and southern California south to
northern Baja California and through the Great Basin
to western and northern Utah and southern Idaho. The
Colorado pinyon pine occurs in Utah, Colorado, Arizona,
and New Mexico.

p. cembroides

Figure 1 — Geographic ranges of Strobus pines
with large, wingless seeds in the Western United
States and Canada.

221

In addition, Parry pinyon (Pinus quadrifolia) and
Mexican pinyon pine {Pinus cembroides), which are lower
elevation species, may occasionally be dispersed by
Clark's nutcracker where their ranges overlap with the
bird, but this has not been documented. Similarly, where
the range of southwestern white pine (Pinus strobiformis,
Subsection Strobi), overlaps with the nutcracker in south-
ern Colorado, New Mexico, and Arizona, its seeds may be
nutcracker-dispersed (see, for example, Benkman and oth-
ers 1984); the ranges of both Mexican pinyon pine and
southwestern white pine continue south into Mexico.

CONE AND SEED TRAITS OF PINES

Large, wingless seeds characterize the four Strobus
pines routinely dispersed by Clark's nutcracker and
southwestern white pine (table 1). In fact, the mean seed
weights of all the wingless-seed Strobus pines that may
be dispersed by Clark's and Eurasian nutcrackers (Nuci-
fraga caryocatactes) (x = 0.279) are significantly greater
than the mean seed weights of wind-dispersed Pinus
pines ix = 0.094) (Wilcoxon test, H = 15.93, P = 0.00006)
and also significantly greater than the weights for the
winged, wind-dispersed Strobus pines (x = 0.057; H = 9.31,
P = 0.0023, Tomback and Linhart 1990).

The cones of whitebark pine are indehiscent and thus
do not open after seeds have ripened (see, for example,
Lanner 1982). In the pinyon pines, the cones open, but
ripe seeds are held in depressions on cone scales by means
of small flanges (see, for example, Vander Wall and Balda
1977). Both seed winglessness and seed retention in
cones increase the chances that nutcrackers rather than
wind will disperse the seeds (Tomback and Linhart 1990).
In contrast, the cones of limber and southwestern white
pine open when the seeds are ripe; however, some limber
pine seeds may remain lodged in open cones as long as
several weeks after ripening (Tomback 1988; Torick
1993). Discussions concerning the evolution of cone and
seed traits of bird-dispersed pines may be found in Lanner
(1980), Tomback (1983), and Tomback and Linhart (1990).

The cones of whitebark pine (Tomback 1978), pinyon
pine (Vander Wall and Balda 1977), and limber pine
(Benkman and others 1984; Tomback and Kramer 1980)
ripen asynchronously within and among trees in late smn-
mer, providing a continuous seed supply for several weeks
or more and allowing nutcrackers to disperse more of the
seed crop. In contrast, the cones of southwestern white
pine ripen synchronously both within and among trees,

Table 1 — Mean seed weights of western Strobus pines
dispersed by Clark's nutcracker. Data from
table 3 in Tomback and Linhart 1 990

Pine species Mean weight

Grams

Singleleaf pinyon (Pinus monophylla) 0.4086
Colorado pinyon (P. edulis) .2387
Whitebark (P. albicaulis) . 1 745

Southwestern white (P. strobiformis) .1 680

Limber (P. flexilis) .0926

Table 2 — Maximum sublingual pouch capacities of Clark's nut-
cracker for four Strobus pines. Scientific names of pines
as in table 1

Maximum number

Pine species

of seeds

Reference

Whitebark

100-150

Tomback 1982; Tomback

and Knowles 1 989

Limber

120

Vander Wall 1988

Colorado pinyon

Vander Wall and Balda

1977

Singleleaf pinyon

Vander Wall 1988

which may result in the dispersal of a smaller percentage
of the seed crop (Benkman and others 1984).

SEED HARVESTING AND CACHING
BY NUTCRACKERS

The morphology and behavior of the Clark's nutcracker
are similar to those of the Eurasian nutcracker. The
Clark's nutcracker has a long, sharp bill that is used to
stab into and loosen closed cone scales or to remove seeds
from open cones (see Tomback 1978 for details of seed
harvesting behavior). Nutcrackers transport pine seeds
in the sublingual pouch, a sac that is formed by the floor
of the mouth and opens under the tongue (Bock and oth-
ers 1973). Maximum seed capacities of the pouch vary
with the pine species (table 2). In addition, nutcrackers
have a remarkable spatial memory that enables them
to relocate their seed stores accurately for more than
9 months after caching them (Balda and Kamil 1992;
Kamil and Balda 1985; Tomback 1980; Vander Wall
1982).

Clark's nutcrackers may begin eating imripe seeds
from the cones of whitebark (Hutchins and Lanner 1982;
Tomback 1978) and limber pine (Tomback and Taylor
1987) as early as mid to late July in some years. A bird
exposes the unripe seeds by stabbing its bill between cone
scales and tearing off scale pieces. Usually only bits of
unripe seeds can be extracted at this time. Nutcracker-
damaged cones increase in frequency throughout summer.
In limber pine they are particularly noticeable: where
nutcrackers removed seeds, red-brown patches of frayed
cone material contrast against the green, undamaged
cone scales.

When seeds are ripe and extracted from cones at a rela-
tively rapid rate (Hutchins and Lanner 1982; Tomback
1978; Vander Wall 1988), nutcrackers begin to cache
them. Nutcrackers have stored whitebark pine seeds
as early as August 15 in the Rocky Mountains (Hutchins
and Lanner 1982) and August 25 in the Sierra Nevada
(Tomback 1978). For limber pine and both pinyon pines,
dates of first seed storage are similar in different montane
areas: late August to early September (Tomback and
Kramer 1980; Tomback and Taylor 1987; Vander Wall
1988; Vander Wall and Balda 1977).

Nutcrackers cache seeds at distances from a few meters
to 22 km from parent trees. Vander Wall and Balda (1977)

222

determined that some nutcrackers carried Colorado pin-
yon seeds 7.5 to 22 km to caching areas, and Tomback
(1978) determined that nutcrackers traveled as far as
8 to 12 km with whitebark pine seeds. More t3T)ical dis-
tances are shorter and up to a few kilometers (Hutchins
and Lanner 1982; Tomback 1978; Vander Wall 1988).

In addition to traveling moderate to long distances,
nutcrackers may fly to higher or lower elevations to store
seeds. For example, nutcrackers in the eastern Sierra
Nevada cache whitebark pine seeds not only at subalpine
elevations (2,700 to 3,000 m) but also at 2,100 m to 2,400 m
in the mixed Jeffrey pine (Pinus jeffreyi) and singleleaf
pinyon pine forest (Tomback 1978). Similarly, nutcrack-
ers carrying Colorado pinyon seeds in northern Arizona
may fly up 300 m elevation to caching areas (Vander Wall
and Balda 1977). In the Colorado Front Range, nutcrack-
ers were observed to carry bristlecone pine {Pinus aristata)
seeds from the subalpine forest up several hvmdred meters
toward tundra for caching (Baud 1993).

Clark's nutcrackers cache seeds both in communal
storage areas, often characterized by steep, south-facing
slopes (Tomback 1978; Vander Wall and Balda 1977), and
in terrain not far from parent trees (Tomback 1978). The
Eurasian nutcracker, in contrast, caches many of its seeds
in a year-round territory (see, for example Mattes 1982).
Seeds are usually buried 1 to 3 cm imder gravelly soil,
mineral soil, or duff (Hutchins and Lanner 1982; Tomback
1978), although a small number may be hidden in hol-
lows, under bark, in cracks, and in holes in trees and in
logs (Tomback 1978; Torick 1993). Each cache typically
consists of one to 15 seeds, although larger caches may
occasionally be made; mean cache sizes are three or four
seeds (Tomback and Linhart 1990, table 5 and references
therein; Torick 1993).

The specific kinds of sites selected by nutcrackers for
biuying seed caches include the following: near the base
of trees; around tree roots, fallen trees, and branches;
around rocks and boulders; on rocky, exposed ledges;
within sparse to heavy plant cover and moss; at the edge
of grassy meadows; and within stands of krummholz pine
forms (Hutchins and Lannner 1982; Tomback 1978, 1986;
Tomback and Kramer 1980; Tomback and Taylor 1987;
Vander Wall and Balda 1977). In addition, nutcrackers
bury seeds at and above treeline (Baud 1993) and in
burned forest terrain (Tomback 1986; Tomback and
Knowles 1989).

Estimates of how many seeds are stored by a single
Clark's nutcracker vary geographically and with the pine
species. For whitebark pine, Tomback (1982) estimated
that about 35,000 seeds per individual were cached after
a good cone crop in the eastern Sierra Nevada, but the
nutcracker then cached the seeds of other pine species,
such as Jeffrey pine or singleleaf pinyon pine, or even lim-
ber pine (Tomback and Kramer 1980). Hutchins and
Lanner (1982) estimated that a single nutcracker cached
as many as 98,000 whitebark pine seeds in western
Wyoming. According to Vander Wall and Balda (1977),
one nutcracker in northern Arizona stored between 22,000
and 33,000 Colorado pinyon pine seeds following a good
cone crop. For northern Utah, Vander Wall (1988) calcu-
lated that one nutcracker stored a maximvun of 16,000

limber pine seeds and a maximimi of about 18,000 single-
leaf pinyon seeds.

In all cases the stored seeds provided several times the
amoimt of energy required by one nutcracker during the
period that the seed stores were used (see, for example,
Tomback 1982; Vander Wall and Balda 1977). The excess
stored seeds probably provided a margin for loss of caches
to rodents, feeding caches to young, and some "forgetting"
of seed cache location (see, for example, Balda and Kamil
1992; Tomback 1982).

PHENOLOGY OF USE AND SPECIES
PREFERENCE

In much of the montane Western United States and
southern Canada, more than one nutcracker-dispersed
Strobus pine occurs in the same geographic area (fig. 1).
For example, in the Northern Rocky Moimtains whitebark
pine t3T)ically occurs at subalpine elevations and limber
pine at the lower forest boundary, although in some loca-
tions the two species may co-occur in the subalpine zone.
In years when both species produce cones, nutcrackers
preferentially harvest and store whitebark pine seeds
before moving into limber pine forests (Tomback 1992).
Similarly, in the eastern Sierra Nevada, whitebark pine,
limber pine, and singleleaf pinyon occur in the same re-
gion; whitebark pine seeds are harvested and stored
before limber pine and singleleaf pinyon pine seeds
(Tomback 1978; Tomback and Kramer 1980). In both
regions, nutcrackers may prefer to take whitebark pine
seeds first, because the cones ripen earlier than do limber
pine or pinyon pine cones.

In the Raft River Mountains of northern Utah, stands
of limber pine and singleleaf pinyon pine occur only a few
kilometers apart. When cone crops of both species were
available, most nutcrackers harvested and stored limber
pine seeds before using pinyon pine seeds (Vander Wall
1988). Limber pine cones ripened before pinyon pine
cones in this area. When limber pine and Colorado pin-
yon pine both produced cone crops in northern Arizona,
Vander Wall and Balda (1977) observed nutcrackers har-
vesting and caching seeds from the closed, green cones of
both species at the same time in late summer. These ex-
amples suggest that species preference is probably based
on cone ripening sequence.

Clark's nutcrackers are also known to harvest and
store the seeds of several wind-dispersed pine species of
both the subgenus Strobus and Pinus. Stomach contents
analysis of Clark's nutcrackers collected in western
Montana indicated that the birds ate, if not cached, pon-
derosa pine (Pinus ponderosa) seeds and Douglas-fir
(Pseudotsuga menziesii) seeds in addition to whitebark
pine seeds (Giimtoli and Mewaldt 1978). Nutcrackers
regularly harvest and cache seeds of Jeffrey pine in the
eastern Sierra Nevada (Tomback 1978). Lanner (1988)
noted nutcracker-damaged cones of Great Basin bristle-
cone pine {Pinus longaeva) in the Wasatch Range and
in the Great Basin. In the Front Range of the Colorado
Rockies, Torick (1993) observed nutcrackers harvest and
cache ponderosa pine seeds, and Baud (1993) observed
nutcrackers harvest and transport bristlecone pine seeds

223

to caching areas. Baud (1993) recently obtained experi-
mental evidence that both bristlecone pine and ponderosa
pine seedlings may be routinely established from nut-
cracker caches. This has been indirectly confirmed by
Torick (1993) who has shown that the tree cluster growth
form, which is foimd in several nutcracker-dispersed
pines and arises from multiseed nutcracker caches
(Tomback and Schuster, these proceedings, £md references
therein), also occurs in ponderosa and bristlecone pine.

ACKNOWLEDGMENTS

Support to the author for attending the SSPE Work-
shop was provided by the College of Liberal Arts and Sci-
ences, University of Colorado at Denver. I thank Lisa
Torick and Karen Baud for comments on an earlier draft
of this paper.

REFERENCES

American Ornithologists' Union. 1983. Check-list of North
American birds. 6th ed. Lawrence, KS: American Orni-
thologists' Union. 877 p.

Balda, R. P.; Kamil, A. C. 1992. Long-term spatial memory
in Clark's nutcracker, Nucifraga columhiana. Animal
Behavior. 44: 761-769.

Baud, K. S. 1993. Germination rates and losses to ro-
dents of artificial nutcracker caches of three species of
Pinus placed in three elevational zones in the Colorado
Front Range. Data on file. Department of Biology, Uni-
versity of Colorado at Denver.

Benkman, C. W.; Balda, R. P.; Smith, C. C. 1984. Adapta-
tions for seed dispersal and the compromises due to
seed predation in limber pine. Ecology. 65: 632-642.

Bock, W. J.; Balda, R. P.; Vander Wall, S. B. 1973. Mor-
phology of the sublingual pouch and tongue muscula-
ture in Clark's nutcracker. Auk. 90: 491-519.

Critchfield, W. F.; Little, E. L., Jr. 1966. Geographic
distribution of the pines of the world. Misc. Publ. 991.
Washington, DC: U.S. Department of Agriculture, For-
est Service. 97 p.

Giuntoli, M.; Mewaldt, L. R. 1978. Stomach contents
of Clark's nutcracker collected in western Montana.
Auk. 95: 595-598.

Hutchins, H. E.; Lanner, R. M. 1982. The central role of
Clark's nutcracker in the dispersal and establishment
of whitebark pine. Oecologia. 55: 192-201.

Kamil, A. C; Balda, R. P. 1985. Cache recovery and spa-
tial memory in Clark's nutcrackers (Nucifraga Colum-
biana). Journal of Experimental Psychology: Animal
Behavior Processes. 11: 95-111.

Lanner, R. M. 1980. Avian seed dispersal as a factor
in the ecology and evolution of limber £ind whitebark
pines. In: Dancik, B. P.; Higginbotham, K. O., eds. Sixth
North American forest biology workshop: proceedings.
Edmonton, AB: University of Alberta: 15-48.

Lanner, R. M. 1982. Adaptations of whitebark pine for
seed dispersal by Clark's nutcracker. Canadian Journal
of Forest Research. 12: 391-402.

Lanner, R. M. 1988. Dependence of Great Basin bristle-
cone pine on Clark's nutcracker for regeneration at high
elevations. Arctic and Alpine Research. 20: 358-362.

Ligon, J. D. 1978. Reproductive interdependence of
pinon jays and pinon pines. Ecological Monographs.
48: 111-126.

Mattes, H. 1982. Die Lebengemeinschaft von Tannenhaher
imd Arve. Birmensdorf, Switzerland: Swiss Federal In-
stitute of Forestry Research. 74 p.

Tomback, D. F. 1978. Foraging strategies of Clark's nut-
cracker. Living Bird. 16: 123-161.

Tomback, D. F. 1980. How nutcrackers find their seed
stores. Condor. 82: 10-19.

Tomback, D. F. 1982. Dispersal of whitebark pine seeds
by Clark's nutcracker: a mutualism hypothesis. Journal
of Animal Ecology. 51: 451-467.

Tomback, D. F. 1983. Nutcrackers and pines: coevolution
or coadaptation? In: Nitecki, M. H., ed. Coevolution.
Chicago: University of Chicago Press: 179-223.

Tomback, D. F. 1986. Post-fire regeneration of krummholz
whitebark pine: a consequence of nutcracker seed cach-
ing. Madrono. 33: 100-110.

Tomback, D. F. 1988. Nutcracker-pine mutualisms: multi-
trunk trees and seed size. In: Ouellet, H., ed. Acta XIX
Congressus Internationalis Ornithologici, Vol. I; 1986
June 22-29. Ottawa, ON: University of Ottawa Press:
518-527.

Tomback, D. F. 1992. Observations in the Wind River
Range. Unpublished data on file, Department of Biol-
ogy, University of Colorado at Denver.

Tomback, D. F.; Knowles, J. W. 1989. Post-fire whitebark
pine seed dispersal by Clark's nutcracker in Yellowstone
National Park. Unpublished data on file, Department
of Biology, University of Colorado at Denver.

Tomback, D. F.; Kramer, K. A. 1980. Limber pine seed
harvest by Clark's nutcracker in the Sierra Nevada:
timing and foraging behavior. Condor. 82: 467-468.

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird-
dispersed pines. Evolutionary Ecology. 4: 185-219.

Tomback, D. F.; Taylor, C. L. 1987. Tourist impact on
Clark's nutcracker foraging activities in Rocky Moim-
tain National Park. In: Singer, F., ed. Wildlife manage-
ment and habitats. Washington, DC: National Park
Service and George Wright Society: 158-172.

Torick, L. L. 1993. Field and laboratory evidence for
regeneration of wind-dispersed pines from animal-
dispersed caches. Data on file, Department of Biology,
University of Colorado at Denver.

Vander Wall, S. B. 1982. An experimental analysis of
cache recovery in Clark's nutcracker. Animal Behavior.
30: 84-94.

Vander Wall, S. B. 1988. Foraging of Clark's nutcrackers
on rapidly changing pine seed resources. Condor. 90:
621-631.

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations
of Clark's nutcracker and the pinon pine for efficient
seed harvest and dispersal. Ecological Monographs.
47: 89-111.

224

Forest Structure
and Dynamics

International Workshop
St. Moritz 1992

STRUCTURE OF SWISS STONE PINE
STANDS IN NORTHEASTERN ITALY

Giovanna De Mas
Elena Piutti

Abstract — This contribution suggests a structural typology for
Swiss stone pine {Pinus cembra) stands in the eastern range of
the ItaUan Alps. Study of the structure in these stands makes it
possible to understand their dynamics and to evaluate environ-
mental and scenic aspects of the present expansion of high-
elevation forests. This paper describes eight structural types
that show different rhythms in the forest establishment process.

The structure of Swiss stone pine {Pinus cembra L.)
stands has been studied by several authors, although with
different methodological approaches (see, for examples,
Contini and Lavarelo 1981, 1982; Del Favero and others
1985; Kuoch 1972; Mayer and Ott 1991; Piussi and
Schneider 1985; Rachoy 1976; Schiechtl and Stern 1975,
1979, 1983, 1984; Stern and Helm 1979). These studies
are of particular interest. They allow one to make an
improved assessment of management and environmental
implications resulting from reforestation of high-elevation
abandoned pastures. These have become a typical feature
of many gdpine areas as a consequence of changed economic
structures.

In fact, since the structure analysis of a tree stand pro-
vides a detailed description of the spatial and temporal
tree distribution pattern, these early approaches are nec-
essary to understand the complex mechanisms that con-
trol natural regeneration and intra- and interspecific com-
petition. Moreover, they enable us to quantify rhythms of
different tree species establishment and to define the com-
plex stand-stability degree. The objective of this investi-
gation, which is part of the INTEGRAL? project finan-
cially supported by EEC, is to detect early structure types
of stone pine stands on the southern slope of the eastern
Alps. Knowledge of these stands is comparatively scanty.

FIELD SURVEYS

Field work was carried out in a large part of the autono-
mous Province of Bolzano and in the Province of Belluno.
In particular, the study was conducted in stone pine

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

Giovanna De Mas is a Researcher and Elena Piutti is making her re-
search doctorate stage. Land and Agro-Forest Environment Department,
University of Padova, 35100 Padova, Italy.

coenoses in the following valleys of Bolzano Province:
Venosta, Passiria, Sarentino, Funes, Luson, Gardena,
and Badia and in Cortina basin in Belluno Province. This
scenic area exhibits a remarkable climatic and geologic
variability.

Climate becomes increasingly continental while mov-
ing fi*om east to west (Del Favero and others 1985; Fliri
1975). Temperatures are moderate in the east and an-
nual precipitation exceeds 1,000 mm, with rainfall max-
ima in spring and autumn. In the west, climate gradually
becomes colder in winter, and summers are dry and hot.
Annual precipitation is less than 1,000 mm and the rain-
fall regime is continental. Geological substrata cheinge
fi'om limestone in the east to silicate bedrock in the west.

In the eastern part, stone pine is found at the eastern
periphery of its natural range. It forms pure or mixed for-
ests associated with larch (Larix decidua Mill.) and/or
spruce (Picea abies Karst.). It is occasionally mixed with
Scots pine (Pinus sylvestris L.) or with silver fir (Abies
alba Mill.). The vegetational aspects of these associations
were described by Filipello and others (1976; 1980-81).

Based on the results of field surveys and on data avail-
able from forest management plans, it was possible to
make a preliminary early subdivision of the stone pine
stands examined into comparatively homogeneous struc-
ture categories. These categories range fi*om open stands
of remarkably large-size trees, sometimes accompanied by
recent young trees, to more or less closed forests.

First, a rough classification was conceived. Then, some
forest stands were subjectively identified by size and vari-
ability of the category itself These stands were consid-
ered to approximately represent the specific structure
conditions. In each of these stands, a rectangular 400-m'^
sample plot was marked. In this plot, the following pa-
rameters were recorded for each tree exceeding a diam-
eter of 7.5 cm at ground level: breast height diameter,
height, age at tree base, distance of tree center fi"om the
two plot vertices. Then, spatial tree distribution (vertical
and horizontal) was drawn with early processing and the
following stand parameters were computed per hectare:
niunber of trees, basal area, voliune (determined using
the double entry volvune tables of the Nationsd Forest
Inventory, I.S.A.F.A., 1984), average diameter, and aver-
age height.

In total, 39 sample plots were surveyed. Their main
characteristics are summarized in table 1. Subjective
sampling and the selection of a comparatively low number
of small sample plots was inevitable because of limited
funding and in view of the objective of analyzing at least
the most important stands among those observed.

226

° 12 ffl
o o S

< o

'x o

> «
< =6

3
O
>

o
o

E
o
o

3
i_

o .£

_ «

(0 £

(0 m

» i

O €

. V)

o
o

Q.
(A
<

>
o

Q- 3

■2

a5TtcoLn-i--^oot^(r)OLntnino^-cooinooooir)OP5t^or3cr)LOLr>LnoooooT-rs.
r~~.cO'r-o5Qot^r^-<tcor^cvjcor>»t>~.ocoooino)oot^-t-coi^O'«*co<oo<35'a-oin-r-c\j-i-i^coo

■"-■^CM •t-i--.-OJCVJ-.-CVJC\Ji-i-i-i-T-T-T-CM't-CJi-'>-^C\JC0t--t- t-C\jT-C\JCVJCVJC\Ji-CVJ

IT)

Lf)

IT)

ir>

cri

IT)

iri

C\J

CVi

OO-i-CMLnCMQOh-O

TtcjjcdcDCMooT-^CMin

T-Lncoqqqi--r..^in^
o'cicoinTtodi^ihT^csco

CO CO C33 in o

CO CM CD CM CM T-^

cDi-coor^coinc7)ooincDCM

T-'cDT-^obodcOCMOOCMCjicjiod

CO O)

<J>

CM
CM

CO
CM

00
CM

CM
CM

00
CM

CO
CM

oci

CD
CM

00
CM

T—

C3>

CJ)

Cli

N.'

iri

■>*■

Cvj

CJ)

CO 1-

CJ) o
in in CO

in CM

CO CM

CD
CM

in <3) in
^ m r-.

CO O)
CD CO

m o CO

00
CO

O 00 CM

in t- CO

o
in

CO T- CO
CM

CM CO CO in

Ti- CM r-- CO

CO CM O
CO CM

IT)

<J)

l>.

C7>

c:^

CT)

■t

0C3

iri

•<t

ir>

oinininoinininininin
inr^cMr^int^^cMCMi^cMCM
inincDincMin'*'<tin't'«t

oinooinooinoooinininin
inr^inincMininN-^oor^cMCMCM
■^•<tcDLncoor^oO'-cocooO'<!tcocM

o in

O CM

oininininininoinoo
or^cMr^cMi^r-mr^oo
cDinincDtcom'tcO'^'a-

C/) C/D C/D LU

Z Z Z Z Z lU z

O LU LU O z
CO CO CO CO LU LU

LU ^ W ^ Uj ^ ^
ZCO^ZZZCOZ

^ ^ ^

LULULULUmZZZ
ZZZZZZZZ

^ LU LU

^ z z

Z Z Z LU UJ

■t

CJ)

T—

<3)

CO{OCOCO(0(OOBCOD3

CDCDOOCDCDCQCDU.U.

3 3 3 3

SOSOOOOOSOOOSCOCOCOCOCOCOCOCOCOOO
-l>>>>>>>>C0C0C0C0Q.CLC0C0C0C0C0C0C0OO

T-cMco'^incor~-.ooc3>oi-CMco'<i-incDr^

cooiOi-CMco^incoi^oocj)Oi-cMco^incor^ooo>

i-i-CMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCO

227

METHODS

As mentioned earlier, the structural analysis of a forest
stand enables us to describe its appearance and to under-
stand its dynamics — the patterns and the mechanisms
that have controlled its development until now. The pres-
ent state results from the interaction between natiiral
components and human disturbances.

The structure of a forest coenosis can be illustrated by
means of its characteristics or by "chronological" ele-
ments. The former can be sampled and observed more
easily. They allow a static view of stand conditions that
can be described in detail. For this purpose, the following
elements are considered in general: composition (numeric
and volumetric), percentage of tree nimiber in diameter
classes (usually 5-cm range) and in height classes (of
different width), related to total sum of trees, numeric
indices of spatial distribution, and some dendrometric
parameters.

In this investigation, the elements given were taken
into consideration for each sample plot. However, only
the following ones were used in the assessment of struc-
tural types: numeric composition, diameter classes, rela-
tive height classes related to three layers, and ratio of
variance with mean of niunber of trees found in the four
subplots, each 100 m^ wide, into which the sample plot
had been divided.

The chronological structure of a stand provides a pic-
ture of the stand dynamics and of species establishment.
Thus the chronological aspect is essential for structure
analysis, although it is often neglected because of survey-
ing difficulties. Its identification is made by subdividing
the number of trees, distinguished by tree species, into
chronological classes with an unstandardized range (in
the case considered in this work, a 20 years' lapse rate
was fixed).

To develop an early structural typology for the stands
investigated, we tried to aggregate sample plots into ho-
mogeneous groups. Distance measures were computed
using Minkowski's algorithm (Franceschetti and Provasi
1978) taking the physical and chronological factors into
consideration, after having standardized data and con-
verting them into percentages of the totals.

Because of the unsatisfactory results obtained in this
early processing, another method was tried. Three sets of
sample plots, different in composition, were selected. The
first set (17 sample plots) includes those where stone pine
represents over 80 percent of the species composition; the
second (14 plots) includes those where stone pine is
present by less than 80 percent; spruce prevails, or is the
second species by importance. The third set (eight plots)
comprises those where larch prevails or is the second spe-
cies after stone pine (which is in any case present by less
than 80 percent). Then, the distance measure computa-
tion was repeated in each of these three sets, considering
as factors only the ones concerning the physical structure.

Before discussing the results, it is necessary to state
that no case-by-case assessment could be made of the
modifications that cutting or grazing caused on the struc-
ture of the single sample plots, because they were hardly

noticeable on small-size sireas. It is only possible to be-
lieve that both activities were carried on in the areas in-
vestigated and the dynamism that led to the present
structure was affected by severe cuts made in the past
and by the progressive abandonment of grazing practices.

RESULTS

Five stand structure types (fig. 1) are distinguished in
the first set of sample areas where stone pine is present
by over 80 percent. Three types, however, are based on
one observation only.

The first type (fig. 2, structural type 1.1), includes six
sample plots. Diameter distribution decreases as diam-
eter increases, with trees that reach the 55-cm class. As
to vertical distribution, a remarkable concentration of
heights is foimd in the intermediate layer, although a fair
share occurs in the upper third.

In the second type (fig. 3, structural type 1.2, eight
sample plots), the regeneration process occurred gradu-
ally as tree age ranges from 20 to 220 years. Most of
them are 101 to 120 years old, while a few exceed 140
years. Diameter distribution is different; it is unimodal
with a maximum in the 20-cm class and with a high con-
centration in the 15- to 30-cm classes. Vertical distribu-
tion shows most heights in the upper third.

In the third type (fig. 4, structural type 1.3), the stand
has a reduced niunber of trees per hectare (140 trees/ha),
and trees are missing at the youngest age. The last new
generation occurred 60 to 140 years ago. The height dis-
tribution of trees appears well balanced in the layers,
even if a slight trend to one-storied formation can be no-
ticed.

Contrary to that, in the fourth type (fig. 5, structural
type 1.4), regeneration appears to be more recent because
the new generation comprises trees 40 to 80 years old.
Tree heights concentrate in the lower third just as do the
diameters, which are mainly foimd within the 10- to 20-
cm classes.

The last type (fig. 6, structural type 1.5) represents the
structural conditions of a comparatively rare pure even-
aged stone pine stand where all mature individuals are
over 220 years old while younger trees are missing. Trees
(400/ha) have diameters varying from 30 to 60 cm and
heights concentrated in the third upper layer.

Three types are distinguished in the second set. They
concern sample plots where stone pine is less than 80 per-
cent and spruce is either prevailing or the second species.

In the first type (fig. 7, structural type 2.1), including
five sample plots, diameters range fi-om 10 to 50 cm and
exhibit a bimodal-type pattern with a main maximum in
the 30-cm class and a secondary maximum in the 20-cm
class. Trees are concentrated in the 20- to 35-cm classes.
The high number of trees in the 10-cm class (youngest
individuals) should be emphasized; this reflects well-
established regeneration. As to vertical distribution, a
trend to one-storied stand formation is evident since over
60 percent of the trees are included in the upper third.

From the chronological point of view, a slight imeven-
age structure is apparent. It is characterized by a wide
range of tree ages varjdng from 20 to over 220 years.

228

200 t

150

100

50 ■

18 30 16 15 29 19 4 23 21 32 10 12 26 33 20 25 27

Plot Numbers (table 1)

200 T

150 -

100 -

37 35 13 3 9 14 1 2 31 17 34 36 39 5

Plot Numbers (table 1)

28 8 6 38 7 22 11 24

Plot Numbers (table 1)

Figure 1 — Dendrograms obtained from the measurements of distance among the elements of
physical structure. A, set of stone pine plots; B, set of stone pine and spruce plots; C, set of
stone pine and larch plots.

229

30-

25-

■a
13

21-40 41-60 61-HO 81-IOU 101-120 I2I-I4U MMM) lAMW 181 21X1 201-220 >22\

chronological age classes

I I P. cembra

Figure 2— Structural type 1.1; including plots 1 8,
30, 16, 15, 29, and 19. The main characteristics of
the structural type are illustrated by average values
of physical and chronological elements. In particu-
lar, the upper part of the figure shows tree fre-
quency in diameter classes and the value of the
average diameter. A second graph shows percent-
age of trees related to three layers. These are sub-
divided by maximum height and represent the
vertical type structure. In addition, average height
and maximum height are shown. A third graph dis-
plays the percentage of trees, distinguished by tree
species, into age classes of 20 years' width. Fi-
nally, a list is given of some average dendrometric
parameters that are related to the hectare (number
of subjects, basal area, volume) and the number of
plots where the spatial distribution index is less
than or more than one, respectively. For this last
parameter, which in any case never affected type
formation in a significant way, the average value
was not computed because it often proved to be
highly variable inside the group.

230

average h = 13 m max. h = 17.6 m

relative frequency

30-

T T r T Y Y y Y T Y-

21-40 41-« 61-80 Hl-lOn 1(11-120 12!-l<0 141-l«) 161 IHO 181-2(0 201-221) >221

chronological age classes

I I P. cembra

Figure 3— Structural type 1 .2; including plots 4, 23, 21 ,
32, 10, 12,26, and 33.

231

d
e

10 15 20 25 30 35 40 45 50 55 60 65 70 75
diameter classes (cm)

60-

chronological age classes

P. cembra

Figure 4— Structural type 1 .3; including plot 20.

232

400

10 15 20 25 30 35 40 45 50 55 60 65 70 75
diameter classes (cm)

average h = 12.9 m max. h = 23 m

0 10 20 30 40 50 60 70 80 90

relative frequency

«5-

T T 1 1 r— 1 1 1

21-40 41-60 (ilSi 81-1(10 101-121) 12M40 141-160 lAI-lM} lill-2in 201-Z20 >221

chronological age classes

I I P. cembra

Figure 5— Structural type 1 .4; including plot 25.

233

10 15 20 25 30 35 40 45 50 55 60 65 70 75
diameter classes (cm)

average h = 21 m max. h = 23.5 m

15.8-23.5

M 7.9-15.7

0-7.8

relative frequency

0 10 20 30^ 40 50 60 70 80 90

-1 1 1 1 1 1 r

21-4(1 4I'M) 61-80 81100 101 120 12I-I«I 141-160 161180 IRl-200 2U1-220 >221

No/ha 400
G/ha m 55
V/ha m 494.1
f . spatial index
<1 1
>1 0

chronological age classes

I I P. cembra

Figure 6— Structural type 1 .5; including plot 27.

234

400

350-
300
^ 250-
i 200-

average dbh = 29.4 cm

150
100
50
0

10 15 20 25 30 35 40 45 50 55 60 65 70 75
diameter classes (cm)

average h = 17.3 m max.h = 22.1

0 10 20 30 40 50 60 70 80 90

relative frequency

T T f 1 1 T T-

21-40 41.00 A|.aO 81-1(111 101.120 121 140 I4MM lAI-IM) 181.2(10 201.221) >22l

chronological age classes

P. cembra

L. decidua

P. abies

Figure 7— Structural type 2.1 ; including plots 37, 35,
13, 3, and 9.

235

There is, however, a large quantity of trees in the 121- to
160-year classes, with a maximum in the class of 121 to
140 years. Moreover, larch is conspicuous only in the ad-
vanced age classes (from 120 years and older, with a
maximum in the 201- to 220-year class). In this type, the
average nimaber of trees per hectare amounts to 625,
basal area per hectare is 42.5 m^, while average volimae
per hectare turns out to be 347.2 m^.

The second type (fig. 8, structural type 2.2), including
four sample plots, shows a wider diameter range (from
10 to 70 cm). A trend can be clearly noticed for unimodal
distribution with a maximum in the 40-cm class. A fair
variability can be observed also in the upright tree distri-
bution. In fact, individuals with heights in the upper
layer prevail; nevertheless individuals in the lower third
are well represented.

A slight uneven-age condition can also be noticed in this
second type (ages range from 20 to 220 years), but age dis-
tribution is more regular with 141- to 160-year classes
prevailing. The average values are 369 trees per hectare,
basal area is 35.2 m^, and volume is 329.2 m^.

The last type of the second group (fig. 9, structural type
2.3) includes five sample plots and, in contrast with those
so far examined, it has a decreasing diameter distribution
with a high nimaber of small- or middle-size trees that are
not strictly found in the yoimger chronological classes.
Height distribution is more uniform. There is a larger
one-storied formation, because nearly 90 percent of the
trees form the two upper layers. Moreover, differences in
age are slightly less conspicuous (40 to 200 years). Most
trees are in the 80 to 140 classes. The number of trees
per hectare (670) is comparatively high, while the many
small-size trees lower the value for basal area (21.8 m^)
and for volrnne (172.8 m^) per hectare. The decreased pro-
portion of larch is emphasized and confirms what already
has been stated. Larch less than 80 years is not repre-
sented in this case; instead other species (Scots pine and
silver fir) join in the formation.

In the third set, including areas where stone pine is
represented by less than 80 percent and larch is either
predominant or the second species, it should be noted that
aggregations obtained through distance measures do not
seem to be satisfactory because of extreme stand differ-
ences. Consequently, it was impossible to single out
groups representing average conditions. It may be that
only a subsequent closer investigation with an increased
number of observations will enable us to define a struc-
tural typology for these coenoses. Larch is found either in
the oldest chronological classes or in the young classes,
where structures are more open.

years can be observed. Most of the stsind, however, con-
sists of trees with ages ranging fi-om 20 to 120 years, but
with a peak in the 60- to 80-year class. Thus it can be
supposed that in this case regeneration has occurred
within a comparatively short time, and is now decreasing
gradually. The stands have a good number of trees per
hectare (662). Conversely, the other dendrometric param-
eters (all of them rather low) suggest that biospace will be
saturated chiefly by a growth increase of individuals al-
ready present rather than by additional trees.

Conversely, in the second type the physical and chrono-
logical aspects and the dendrometric parameters, which
are higher than in the previous type (531 trees/ha,
25.6 m^/ha basal area, 179.1 m^ mass-volimie-stock/ha)
suggest that biospace filling processes are more advanced
and that the formation is structurally more stable.

As already stated, the three types that complete the
structural typology of the first set are each represented
by only one observation. They illustrate structural condi-
tions that are not very frequent but clearly distinct from
those examined so far. In particular, the third and fovirth
type distinctly display some possible early stages of the
regeneration process, since two clearly different genera-
tions coexist at the same time. A high proportion of trees
over 220 years of age can be observed in both types, while
the remaining individuals fall within yoimger classes. The
structure of the third type can be easily explained consider-
ing that the plot is located at the treeline (2,260 m). This is
an environment where the regeneration process occurs only
occasionally or it may be entirely missing for a long time.
The latter seems to have happened in this case.

In the fourth type, the large number of trees per hec-
tare (625) suggests that the future structure in this stand
should be similar to the first two types described.

The last type, as already pointed out, represents a
rather rare situation.

In the first type of the second set, the chronological
analysis suggests that the stand originated from a grazed
forest. Spruce, as well as stone pine, is represented in all
classes, which means that there was a continuous inva-
sion of these two species into the formation.

The second type is characterized by formations that are
open grown, permitting new spruce and stone pine to con-
tinuously invade the stand. However, larch regeneration
ended about 80 years ago.

In the last type, regeneration occurred rapidly and a
sufficiently closed stand established in a short time. After
that, regeneration slowed down, but competition among
individuals increased. That caused the diameter differen-
tiation mentioned earlier.

DISCUSSION

For the first two sets of stand structure, homogeneous
structural types can be interpreted from the chronological
point of view. The third set was different.

In the first type, physical elements show a structure
characterized by an intermediate layer that is sufficiently
differentiated by competition and is mixed with remark-
ably large-size trees from a previous stand. Taking into
consideration age distribution, few trees older than 120

CONCLUSIONS

From the methodological point of view, the use of dis-
tance measures has made it possible to point out a struc-
tural typology in stands where stone pine, or spruce with
stone pine, are clearly predominant. These results were
obtained using physical structure elements as discrimi-
nating factors. Other results — more difficult to interpret —
were obtained that take into consideration both physical
and chronological elements.

236

10 15 20 25 30 35 40 45 50 55 60 65 70 75
diameter classes (cm)

T r Y T

21-40 M-W fil-flO SMCn 101.12U I2M 40 141-160 ICiMMU IKi-20Q 2UI-220 >221

chronological age classes

I I P. cembra IB L. decidua P. abies

Figure 8— Structural type 2.2; including plots 5, 34, 36,
and 39.

237

238

The structural typology defined, though rather variable,
shows that the regeneration process started about one
century ago, and it sometimes developed in a gradual and
continuous way and at other times more quickly and mas-
sively. Generally, it is possible to observe good structures
in physical and chronological elements. These will prob-
ably provide these stands with sufficient stability.

REFERENCES

Contini, L.; Lavarelo, Y. 1981. Le pin cembro: repartition,
ecologie et croissance. CNRF, Nancy. 254 p.

Contini, L.; Lavarelo, Y. 1982. Le pin cembro: vegetation,
ecologie, Sylviciilt\ire et production. INRA, Paris. 197 p.

Del Favero, R.; De Mas, G.; Lasen, C; Paiero, P. 1985. II
pino cembro nel Veneto. Assessorato Agric. e For., Dip.
per le foreste e I'economia montana, Venezia. 85 p.

FiUpello, S.; Sartori, F.; Vittadini, M. 1976. Le associazioni
del cembro sul versante meridionale dell'arco alpino. Atti
Istituto Botanica Universita di Pavia. 11: 21-104.

Filipello, S.; Sartori, F.; Vittadini, M. 1980-1981. Le associ-
azioni di cembro nel versante meridionale dell'arco
alpino. 2: La vegetazione: aspetti forestali. Atti Istituto
Botanica Universita di Pavia. 14: 1-48.

Fliri, F. 1975. Das Klima der Alpen im Raimie von Tirol.
Universitaets-Verlag Wagner, Innsbruck-Mxmchen.
454 p.

Franceschetti, G.; Provasi, C. 1978. La cluster analysis
quale strumento di zonizzazione nella pianificazione.
In: Metodologia per la formazione dei piani zonali

agricoH nella Regione Veneto. Istituto Estimo Rxirale

e Contabilita, Universite di Padova: 32-90.
I. S. A. F. A. 1984. Inventaiio forestale nazionale. Tavole

di cubatura a doppia entrata. Trento. Ill p.
Kuoch, R. 1972. Zur Struktur und Behanlimg von

Subalpinen Fichtenwaeldem. Schw. Zeit. fur Forst.

123(2): 77-89.

Mayer, H.; Ott, E. 1991. Gebirgswaldbau Schutzwald-
pflege. 2 Auflage, Gustav Fischer Verlag Stuttgart-
New York.

Piussi, P.; Schneider, A. 1985. I limiti superiori del bosco
e degli alberi in Val di Vizze (prov. di Bolzano). Annali
Acc. It. Sc. Forestali: 121-150.

Rachoy, W. 1976. Waldbauliche Stniktunmterschungen
in subalpinen Zirbenwaeldem. Invited Papers, Con-
gress Group 1, XVI. lUFRO-World-Congress, Oslo 1976:
183-202.

Schiechtl, H. M.; Stem, R. 1975. Die Zirbe iPinus cemhra
L.) in den Ostalpen. I Teil, Angew. Pflanzensoz. 22:
1-83.

Schiechtl, H. M.; Stem, R. 1979. Die Zirbe {Pinus cemhra
L.) in den Ostalpen. II Teil, Angew. Pflanzensoz. 24:
1-79.

Schiechtl, H. M.; Stem, R. 1983. Die Zirbe {Pinus cembra
L.) in den Ostalpen. Ill Teil, Angew. Pflanzensoz. 27:
1-110.

Schiechtl, H. M.; Stem, R. 1984. Die Zirbe (Pinus cembra
L.) in den Ostalpen. IV Teil, Angew. Pflanzensoz. 28:
1-99.

Stem, R.; Helm, G. 1979. Alter und Struktur von
Zirbenwaldem. Allgem. Forstr. 90: 194-198.

239

ROLE OF PINUS PUMILA IN PRIMARY
SUCCESSION ON THE LAVA FLOWS OF
VOLCANOES OF KAMCHATKA

S. Yu. Grishin

Abstract— The succession of the subalpine vegetation was studied
on the lava flows of the Central Kamchatka. The rate of succes-
sion depends on type of lava and the formation of friable substra-
tum on lava surface following ash falls and other causes. Stone
pine (Pinus pumila) is not a pioneer on the original lava surface.
Rather, its main role is determined during the first half of succes-
sion (0 to 1,000 years). During this period it forms closed commu-
nities and soil cover, favoring the establishment of other dominant
species in the subalpine zone. During the second half of succession
(1,000 to 2,000 years) P. pumila decreases to the cover level typi-
cal for climax vegetation. Thxis the role of P. pumila is unique; it
restores vegetation cover on juvenile substrate.

Many thousand square kilometers of Kamchatka's terri-
tory are covered by layers of lava flows. Most of the terri-
tory has subalpine vegetation. Under these conditions the
subalpine stone pine (Pinus pumila) (nomenclature of all
species follows Czerepanov 1981) is the most important spe-
cies forming plant cover on juvenile substrate. This spe-
cies usually forms dense commimities in the subalpine belt

vole. Shiveluch #

Figure 1 — Location of study area.

Paper presented at the International Workshop on Subalpine Stone
Pines and Their Environment: The Status of Our Knowledge, St. Moritz,
Switzerland, September 5-11, 1992.

S. Yu. Grishin is Research Scientist at the Institute of Biology and Pedol-
ogy of Far East Branch of Russian Academy of Sciences, Vladivo8tok-22
690022, Russia.

Figure 2— Tolbachik volcano (altitude 3,682 m).

in northeastern Asia. At Kamchatka, p-la P. pumila may
be found from the seashore to the mountain slopes up to
1,400 m above sea level. It covers 42 percent of the forest-
covered territory.

This study was conducted in the Central Kamchatka,
near Tolbachik volcano (figs. 1 and 2). During the Holocene,
lava flows formed a lava plateau with an area of 875 km'^.
The lava flows of Kluchevskoy volcano were also investi-
gated. The age of the flows was determined by the tephra-
chronology method, based on radiocarbon dating (Braitseva
and others 1981). With this method we could distinguish
the stages of succession and estimate their duration
(Grishin 1992).

The climatic conditions of the Tolbachik area reveal a con-
tinental pattern, similar to the climate of Middle Siberia.
The mountain vegetation is composed by forests of spruce
(Picea ajanensis) and larch (Larix kamtschatica) up to an
altitude of 500 m, while birch (Betula ermanii) occurs up to
900 m. The subalpine zone is situated at the interval of
800 to 1,100 m. It Dhas a complex structure: forest islands
of birch, open woodlands of larch, vast covers of stone pine
and alder (Duschekia kamtschatica) krummholz, patches
of subalpine meadows and alpine-tj^je communities. The
alpine belt extends to 2,000 m altitude, and the snowline
lies near 2,500 m (Grishin 1988a,b).

In the summer of 1975 in the central part of the plateau
there was a major eruption. As a result of this, lava flows
with an area of about 9 km^ were effused and tephra (vol-
canic ash, s£ind, and scoria) was dispersed on 470 km^. This

240

Figure 3 — Volcanic plateau of the Tolbachik area:
lava flows covered by tephra and remnants of larch
trees and Pinus pumila krummholz.

layer has a depth of more than 0.1 m (fig. 3). The vegeta-
tion was completely destroyed on 400 km^ (Fedotov 1984).

ROLE IN SUCCESSION

The investigation of the lava flows (10 to 50 years old)
showed that P. pumila is not a pioneer species actively es-
tablishing on the juvenile substrata of pahoehoe (ropy lava),
aa (scoria lava), and tephra. Herbaceous plants (Chamerion
angustifolium, Poa malacantha, Leymus ajanensis, and
others), mosses (Polytrichum spp.), and lichens (Stereocau-
lon vesuvianum) establish most actively initially. Woody
plants such as poplar {Populus suaveolens), some willows,
and very occasionally larch and stone birch establish less
readily — all the plants settle on the friable substratimi
(tephra, products of weathering of lava, and others). Seed-
lings of P. pumila were rather rare, appearing a few years

Figure 4 — Seedling of Pinus pumila on aa lava.

Figure 5 — Lava flow about 500 years old coverea
by tephra to a depth of about 40 cm.

after the eruption (fig. 4). The difficulties in its distribution
are because its seeds are dispersed by animals and seed
sources 5 to 10 km away. Five himdred years later, scat-
tered herbaceous and shrubby vegetation is formed on lava
surface with single (about 10 trees/ha), extremely stimted
(3 to 5 m height) larch trees. Low thickets of P. pumila
(0.3 to 0.5 m height) dominate the vegetation, covering
about 30 percent of the surface (fig. 5). Pinus pumila fills
all potential habitats with friable substratimi and forces
out the herbaceous pioneer species. Under the tephra de-
posit on blocks and plates of lava, a thin layer (1 to 3 cm) of
primitive soil develops. Pinus pumila is rooted in crevices
between blocks on concentrated finable substratum enriched
with organic matter (fig. 6).

On the 1,000-year-old lava flows, more closed vegetation
can be observed. Pinus pumila is the dominant species and
covers 60 to 90 percent of the surface (fig. 7). Its height

Figure 6 — The remnants of stunted larch and
branches of Pinus pumila between blocks of lava.

241

Figure 7 — The flow of aa lava about 1 ,000 years old.

reaches 1 to 1.5 m. The characteristic feature of this stage
of succession is the appearance of dominants of new commu-
nities. The most important plant in the new communities
is alder krummholz. It settles on the substratum prepared
by P. pumila, which eventually disappears from the slopes
of the lava ridge and occurs only in the flat bottoms between
ridges (fig. 8). Together with stone pine, slowly growing
small larch trees (a few dozen trees/ha) are found. The
abimdant leaf fall of the alder and of the undergrowth plants
(Calamagrostis langsdorffii, and others) considerably accel-
erates the rate of succession and prepares the substratvmi
for stone birch — a very important dominant tree in the sub-
alpine forest. The depth of the soil on different lava flows
is 5 to 15 cm.

During the next 500 years (1,000 to 1,500 years old) the rate
of succession increases noticeably, leading to differentiation
of vegetation in the subcdpine zone. On the 1,500-year-old

lava flows, typical subalpine communities can be distin-
guished. In addition to the dominant P. pumila thickets,
the single clumps of alder, and the open woodlands of larch,
communities of stone birch, covers of alder, subalpine mead-
ows, and fragments of alpine heath appear. At this point,
the coverage of P. pumila decreases, occupying its own
typical habitats such as rock crests, concave hollows with
frozen soils, and wind-exposed sites. The thickness of the
soil profile increases to 20 cm, and the leveling of volcano-
genic mi crorelief begins to occur. Both coenotical and ver-
tical differentiation of vegetation takes place. During the
earlier stage of succession, P. pumila dominated at altitudes
fi'om 200 to 1,000 m, but to this stage the forest belt below
800 m is formed, composed of larch and stone birch.

On the 2,000- and 2,500-year-old lava flows the vegeta-
tion was destroyed completely by the ash fall of 1975 (fig. 9).
Its remnants give evidence that vegetation cover had been
approaching the climax stage. It had mature commxmities
with well- developed trees and krvunmholz, the size and
cover of which were related to their altitudinal position.
For example, at one of the flows at the altitude of 700 m
a well-developed birch-larch subalpine forest with some
spruce was located. The birch trees had a height of more
than 10 m and a diameter at breast height of 25 to 40 cm.
The larch trees were about 20 m and 30 to 40 cm, respec-
tively. The alder krummholz under the forest layer had a
height of 2.5 to 3 m, and P. pumila heights were 1.5 to 2 m.

These examples show that under such conditions, begin-
ning at the age of 1,500 years, the vegetation tends to ap-
proach the climatic climax similar to the neighboring locah-
ties. Many criteria prove it.

The peculiarities of succession are determined mainly
by the pattern of lava cover and the accumulation of fiiable
substratvmi on the lava surface. On pahoehoe lava, estab-
lishment of plants on the smooth monolith surface is an
extremely slow process. After the first 1,000 years of such
lithosere, the lava is commonly covered only by crustaceous
lichens. Higher plants settle on friable substratvun in cracks
and cavities, and that is in fact the beginning of psammo-
sere (fig. 10). Even if these plants are rooted in lava, they

Figure 8 — The distribution of vegetation across aa lava flow about 1 ,000 years old. 1 — alder krummholz;
2 — shrubs of Pinus pumila; 3 — petrophilous apeu of groups in the lava outcrops; 4 — bunches of grasses;
5 — lava matter; 6 — tephra of 1975; 7 — buried snow patches.

242

Figure 9 — Timberline ecotone vegetation killed by ash
fall on a lava flow that was about 2,000 years old.

Bezymyanny were also deposited. The negative influence
of major ash falls on vegetation was shown by the 1975
eruption. Pinus pumila krummholz and subalpine birch
forests were killed under the tephra deposit of more than
20 cm, alder krummholz imder more than 30 cm, and larch
forests imder more than 40 cm. Smaller amounts of tephra
(10 to 15 cm) lead to the death of vegetation in subalpine
meadows and alpine heaths. After moderate ash fall, a sec-
ondary succession is possible, and thus lithosere may be
transformed to psammosere.

There are other essential factors of succession such as cH-
matic fluctuations of forest and subalpine vegetation at their
upper limit and forest fires. Fires may originate from vol-
canic eruption. Communities of coniferous trees (larch,
spruce) and especially krimimholz (P. pumila) with dry
lichens and dwarf shrubs from Ericacaea are only slightly
protected from fires.

Thus, the role of P. pumila in the succession on the lava
flows of the Tolbachik area is different at diff'erent stages
of succession. Because of its ecological and biological char-
acteristics, it is not a pioneer at the initial stage. Pinus
pumila plays its main role during the first half of succes-
sion (up to approximately 1,000 years). It forms closed
commimities and soil cover, and favors the establishment
of dominant species in new communities of the subalpine
zone. During the second half of succession, plant cover of
P. pumila gradually decreases to the level typical of climax
vegetation (fig. 11).

CONCLUDING REMARKS

On the whole, the role of P. pumila in succession is unique:
It restores vegetation cover on juvenile substrate unfavor-
able for the establishment and development of woody and
sluiibby species. This feature of P. pumila is realized espe-
cially in the regions with continental climate. In places with
oceanic cUmate, for example on the slopes of the Kluchevskoy
volcano, the coniferous dominants P. pumila and Larix

Figure 10 — Suppressed Pinus pumila and stunted
trees of larch on pahoehoe lava about 1 ,000 years old.

cannot populate all the surface of the pahoehoe, and the
next stages of succession are dependent on the space corre-
lation between lithosere and psammosere. On aa lava this
correlation changes to psammosere more qvdckly, and that
accelerates succession. The final stages of succession on
different types of lava are probably similar because a thick
soil-tephra profile is gradually formed, smoothing out the
irregularities of lava.

Another factor important to the succession rate is climate.
In the area studied, the duration of succession is two to three
times longer than in the oceanic climate of the Japanese
subtropics, where it continues for only 700 years (Tagawa
1964).

A very important factor following volcanism is ash fall
damage to vegetation. For example, in the Tolbachik area,
about 10 eruptions took place during the last 2,000 years.
Four of them were similar to the eruptions of 1975; in that
period the tephra of volcanoes Shiveluch, Kluchevskoy, and

1,000 1,500

Age (yr)

Figure 11 — Generalized chronocline of succession
on lava flows of the Tolbachik are^. 1 — pioneer
unclosed groups of grasses, mosses, lichens;
2 — comm

Featured

Top

Featured

Top

Featured

Top

Featured

Top

Featured

Top

Full text of "Proceedings : International Workshop on Subalpine Stone Pines and Their Environment: the Status of Our Knowledge, St. Moritz, Switzerland, September 5-11, 1992"

See other formats